Researcher profile

Lei Zhang

Lei Zhang contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate

Trust snapshot

Quick read

Trust 21 - EmergingVerification L1Unclaimed author

246works

0followers

48topics

4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2026arXiv

A Derivative-Free Saddle-search Algorithm With Linear Convergence Rate

We propose a derivative-free saddle-search algorithm designed to locate transition states using only function evaluations. The algorithm employs a nested architecture consisting of an inner eigenvector search and an outer saddle-point search. Through rigorous numerical analysis, we prove the almost sure convergence of the inner step under suitable assumptions. Furthermore, we establish the convergence of the outer search using a decaying step size, while demonstrating linear convergence under constant step size and boundedness conditions. Numerical experiments are provided to validate our theoretical results and demonstrate the algorithm's practical applicability.

preprint2026arXiv

A domain decomposition approach to pore-network modeling of porous media flow

We propose a domain-decomposition pore-network method (DD-PNM) for modeling single-phase Stokes flow in porous media. The method combines the accuracy of finite-element discretizations on body-fitted meshes within pore subdomains with a sparse global coupling enforced through interface unknowns. Local Dirichlet-to-Neumann operators are precomputed from finite-element solutions for each pore subdomain, enabling a global Schur-complement system defined solely on internal interfaces. Rigorous mathematical analysis establishes solvability and discrete mass conservation of the global system. Moreover, we constructively recover classical pore-network models by fitting half-throat conductivities to local Dirichlet-to-Neumann maps, providing a principled bridge between mesh-based and network-based frameworks. Numerical results are presented to demonstrate the validity and effectiveness of the overall methodology.

preprint2026arXiv

A Landau-de Gennes Type Theory for Cholesteric-Helical Smectic-Smectic C* Liquid Crystal Phase Transitions

We present a rigorous mathematical analysis of a modified Landau-de Gennes (LdG) theory modeling temperature-driven phase transitions between cholesteric, helical smectic, and smectic C* phases. This model couples a tensor-valued order parameter (nematic orientational order) with a real-valued order parameter (smectic layer modulation). We establish the existence of energy minimizers of the modified LdG energy in three dimensions, subject to Dirichlet conditions, and rigorously analyze the energy minimizers in two asymptotic limits. First, in the Oseen--Frank limit, we show that the global minimizer strongly converges to a minimizer of the Landau-de Gennes bulk energy. Second, in the limit of dominant elastic constants, we prove that the global minimizers converge to a classical helical director profile. Finally, through stability analysis and bifurcation theory, we derive the complete sequence of symmetry-breaking transitions with decreasing temperature-from the cholesteric phase (with in-plane twist and no layering) to an intermediate helical smectic phase (with in-plane twist and layering), and ultimately to the smectic C* phase (with out-of-plane twist and layering). These theoretical results are supported by numerical simulations.

preprint2026arXiv

Adversarial Defense in Vision-Language Models: An Overview

The widespread use of Vision Language Models (VLMs, e.g. CLIP) has raised concerns about their vulnerability to sophisticated and imperceptible adversarial attacks. These attacks could compromise model performance and system security in cross-modal tasks. To address this challenge, three main defense paradigms have been proposed: Training-time Defense, Test-time Adaptation Defense, and Training-free Defense. Training-time Defense involves modifying the training process, typically through adversarial fine-tuning to improve the robustness to adversarial examples. While effective, this approach requires substantial computational resources and may not generalize across all adversarial attacks. Test-time Adaptation Defense focuses on adapting the model at inference time by updating its parameters to handle unlabeled adversarial examples, offering flexibility but often at the cost of increased complexity and computational overhead. Training-free Defense avoids modifying the model itself, instead focusing on altering the adversarial inputs or their feature embeddings, which enforces input perturbations to mitigate the impact of attacks without additional training. This survey reviews the latest advancements in adversarial defense strategies for VLMs, highlighting the strengths and limitations of such approaches and discussing ongoing challenges in enhancing the robustness of VLMs.

preprint2026arXiv

Amplitude analysis and branching fraction measurement of $J/ψ\to Λ\barΣ^0η+\mathrm{c.c}$

Based on a sample of $(10087\pm44)\times10^{6}$ $J/ψ$ events collected with the BESIII detector, a partial-wave analysis of $ J/ψ\to Λ\bar{ Σ}^0η+\mathrm{c.c} $ is performed for the first time. The dominant contributions are found to be excited $Λ$ states with $J^P=1/2^-$ and $J^P=1/2^+$ in the $ηΛ$ mass spectra. The measured masses and widths are $M=1668.8\pm3.1\pm21.2$ MeV/$c^2$ and $Γ=52.7\pm4.2\pm17.8$ MeV for the $Λ(1670)$, and $M=1881.5\pm16.5\pm20.3$ MeV/$c^2$ and $Γ=82.4\pm18.2\pm8.9$ MeV for the $Λ(1810)$, respectively. The branching fraction is determined to be $ \mathcal{B}(J/ψ\to Λ\bar{ Σ}^0η+\mathrm{c.c}) $ = $(3.44 \pm 0.11 \pm 0.13) \times 10^{-5}$. The first uncertainties are statistical and the second systematic.

preprint2026arXiv

Beyond Continuity: Simulation-free Reconstruction of Discrete Branching Dynamics from Single-cell Snapshots

Inferring cellular trajectories from destructive snapshots is complicated by the challenges of stochasticity and non-conservative mass dynamics such as cell proliferation and apoptosis. Existing unbalanced Optimal Transport (OT) methods treat mass as a continuous fluid, performing inference at the population level. However, this macroscopic view often fails to capture the discrete, jump-like nature of birth-death events at single-cell resolution, which is essential for understanding lineage branching and fate decisions. We present Unbalanced Schrödinger Bridge (USB), a simulation-free framework for learning underlying dynamics that effectively integrates both stochastic and unbalanced effects which also models the discrete, jump-like birth-death dynamics at single-cell resolution. Theoretically, USB provides a tractable solution to the Branching Schrödinger Bridge (BSB) problem, offering a rigorous microscopic interpretation where individual cells undergo both Brownian motion and discrete birth-death jumps. Technically, the method implements an efficient solver by introducing a simulation-free training objective that effectively scales to high-dimensional omics data. Empirically, we demonstrate on both simulated and real-world datasets that USB not only achieves trajectory reconstruction performance better than or comparable to deterministic baselines but also uniquely enables realistic discrete simulation of birth-death dynamics at single-cell resolution.

preprint2026arXiv

Breaking Modality Heterogeneity in Low-Bit Quantization for Large Vision-Language Models

Low-bit post-training quantization (PTQ) is a pivotal technique for deploying Vision-Language Models (VLMs) on resource-constrained devices. However, existing PTQ methods often degrade VLMs' accuracy due to the heterogeneous activation distributions of text and vision modalities during quantization. We find that this cross-modal heterogeneity is distributed unevenly across channels: a small subset of channels contains most modality-specific outliers, and these outliers typically reside in different channels for each modality. Motivated by this, we propose SplitQ, a channel-Splitting-driven post-training Quantization framework. At its core, SplitQ introduces a novel Modality-specific Outlier Channel Decoupling (MOCD) module that effectively isolates salient modality-specific outlier channels with minimal overhead. To further address the remaining cross-modal distribution discrepancies, we design an Adaptive Cross-Modal Calibration (ACC) module that employs dual lightweight learnable branches to dynamically mitigate modality-induced quantization errors. Extensive experiments on popular VLMs demonstrate that SplitQ significantly outperforms existing approaches across 6 popular multi-modal datasets under all evaluated quantization settings, including W4A8, W4A4, W3A3, and W3A2. Notably, SplitQ preserves 93.5% of FP16 performance under the challenging W3A3 setting (69.5 vs. 74.3), pushing the efficiency frontier for deploying advanced VLMs. Our code is available at https://github.com/EMVision-NK/SplitQ

preprint2026arXiv

Characterization of FBK NUV-HD-Cryo SiPMs near LHe temperature

Five FBK ``NUV-HD-Cryo'' SiPMs have been characterized at 7 K and 10 K, with 405 nm and 530 nm LED light, respectively. The dark count rate (DCR) was measured to be $\sim$ 1 Hz for the $\sim$ 100 mm$^2$-size SiPMs, or 0.01 Hz/mm$^2$, which is $\sim$ 7 orders lower than the DCR at room temperature (RT). Given the very low DCR at these cryogenic temperatures, we measured the SiPMs' I-V curves with such a method: illuminated the SiPMs with weak light, which differs from the conventional measurements at RT. Then, we measured the photo-detection efficiency (PDE), after-pulse (AP), and cross-talk (CT) with a bias voltage ranging from overvoltage (OV) 5 to 11 V. At the OV interval (5 to 11 V), the PDE was between 20\% - 45\%, and the AP and CT were both between $\sim$ 5\% and $\sim$ 20\%. With an OV higher than 10 V, the PDE would be $\ge$ 40\%, and the AP and CT are $\sim$ 20\%. Combining all of the measurements, we are confident that the SiPMs can be equipped as the photosensors on liquid helium detectors, including but not limited to the time projection chambers, which we have proposed in hunting for low-mass dark matter directly and beyond.

preprint2026arXiv

CMTA: Leveraging Cross-Modal Temporal Artifacts for Generalizable AI-Generated Video Detection

The proliferation of advanced AI video synthesis techniques poses an unprecedented challenge to digital video authenticity. Existing AI-generated video (AIGV) detection methods primarily focus on uni-modal or spatiotemporal artifacts, but they overlook the rich cues within the visual-textual cross-modal space, especially the temporal stability of semantic alignment. In this work, we identify a distinctive fingerprint in AIGVs, termed cross-modal temporal artifact (CMTA). Unlike real videos that exhibit natural temporal fluctuations in cross-modal alignment due to semantic variations, AIGVs display unnaturally stable semantic trajectories governed by given input prompts. To bridge this gap, we propose the CMTA framework, a cross-modal detection approach that captures these unique temporal artifacts through joint cross-modal embedding and multi-grained temporal modeling. Specifically, CMTA leverages BLIP to generate frame-level image captions and utilizes CLIP to extract corresponding visual-textual representations. A coarse-grained temporal modeling branch is then designed to characterize temporal fluctuations in cross-modal alignment with a GRU. In parallel, a fine-grained branch is constructed to capture intricate inter-frame variations from integrated visual-textual features with a Transformer encoder. Extensive experiments on 40 subsets across four large-scale datasets, including GenVideo, EvalCrafter, VideoPhy, and VidProM, validate that our approach sets a new state-of-the-art while exhibiting superior cross-generator generalization. Code and models of CMTA will be released at https://github.com/hwang-cs-ime/CMTA

preprint2026arXiv

Complex Monge-Ampère equation in Orlicz space and Diameter Bound

In this paper, we establish diameter bounds for compact Kähler manifolds equipped with Kähler metrics $ω$, assuming the associated measure lies in a specific Orlicz space and satisfies an integrability condition. Firstly, we prove a priori estimates for solutions of the complex Monge-Ampère equation in Orlicz spaces, encompassing $L^{\infty}$ and stability estimates. This is achieved by employing Kołodziej's approach \cite{Ko98} and the argument of Guo-Phong-Tong-Wang \cite{GuPhToWa21}, respectively. Secondly, building on the work of Guo-Phong-Song-Sturm \cite{GuPhSoSt24-1}, we derive the uniform (local/global) estimates of the Green's function and its gradient for the associated Kähler metric $ω$.

preprint2026arXiv

Cross section measurement of $e^{+}e^{-}\rightarrow π^{0}π^{0}ψ(3686)$ from $\sqrt{s}=$ 4.008 GeV to 4.951 GeV

Using data samples with a total integrated luminosity of $22.1~\rm fb^{-1}$ at center-of-mass energies between 4.008 and 4.951~GeV collected with the BESIII detector, the cross sections of $e^{+}e^{-}\rightarrow π^{0}π^{0}ψ(3686)$ process are measured. The obtained cross sections are found to be approximately one-half of those of $e^{+}e^{-}\rightarrow π^{+}π^{-}ψ(3686)$, consistent with the isospin symmetry expectation. A coherent fit to the dressed cross sections is performed, with the $Y(4230)$~parameters fixed at the values measured in $e^{+}e^{-}\rightarrow π^{+}π^{-}ψ(3686)$. The significances of the $Y(4390)$ and $Y(4660)$ are both larger than $5σ$, and their masses and widths are consistent with the previous measurement in the $e^{+}e^{-}\rightarrow π^{+}π^{-}ψ(3686)$ process.

preprint2026arXiv

D$^3$R-DETR: DETR with Dual-Domain Density Refinement for Tiny Object Detection in Aerial Images

Detecting tiny objects plays a vital role in remote sensing intelligent interpretation, as these objects often carry critical information for downstream applications. However, due to the extremely limited pixel information and significant variations in object density, mainstream Transformer-based detectors often suffer from slow convergence and inaccurate query-object matching. To address these challenges, we propose D$^3$R-DETR, a novel DETR-based detector with Dual-Domain Density Refinement. By fusing spatial and frequency domain information, our method refines low-level feature maps and utilizes their rich details to predict more accurate object density map, thereby guiding the model to precisely localize tiny objects. Extensive experiments on the AI-TOD-v2 dataset demonstrate that D$^3$R-DETR outperforms existing state-of-the-art detectors for tiny object detection.

preprint2026arXiv

Detecting AI-Generated Images via Distributional Deviations from Real Images

The rapid advancement of generative models has significantly enhanced the quality of AI-generated images, raising concerns about misinformation and the erosion of public trust. Detecting AI-generated images has thus become a critical challenge, particularly in terms of generalizing to unseen generative models. Existing methods using frozen pre-trained CLIP models show promise in generalization but treat the image encoder as a basic feature extractor, failing to fully exploit its potential. In this paper, we perform an in-depth analysis of the frozen CLIP image encoder (CLIP-ViT), revealing that it effectively clusters real images in a high-level, abstract feature space. However, it does not truly possess the ability to distinguish between real and AI-generated images. Based on this analysis, we propose a Masking-based Pre-trained model Fine-Tuning (MPFT) strategy, which introduces a Texture-Aware Masking (TAM) mechanism to mask textured areas containing generative model-specific patterns during fine-tuning. This approach compels CLIP-ViT to attend to the "distributional deviations"from authentic images for AI-generated image detection, thereby achieving enhanced generalization performance. Extensive experiments on the GenImage and UniversalFakeDetect datasets demonstrate that our method, fine-tuned with only a minimal number of images, significantly outperforms existing approaches, achieving up to 98.2% and 94.6% average accuracy on the two datasets, respectively.

preprint2026arXiv

First Measurement of the Absolute Branching Fraction of $η_c \to γγ$

We apply a tag-and-probe method to precisely measure the absolute branching fraction of the decay $η_c \to γγ$ with the BESIII experiment at BEPCII. Starting with a large initial sample of $2712.4\pm 14.3$ million $ψ(3686)$ events, a sample of 0.16 million $η_c$ events are tagged via the golden channel $ψ(3686)\to π^0 h_c$, $h_c\to γη_c$, effectively avoiding interference effects. The absolute branching fraction of $η_c \to γγ$ is measured for the first time to be $\mathcal{B}(η_c \to γγ) = (2.45 \pm 0.48_{\rm stat.} \pm 0.09_{\rm syst.}) \times 10^{-4}$. Using the world average value of the total width of the $η_c$, the partial decay width of $η_c \to γγ$ is calculated to be $Γ(η_c \to γγ) = (7.48 \pm 1.48_{\rm stat.} \pm 0.30_{\rm syst.})~{\rm keV}$.

preprint2026arXiv

First Observation of $D^{0(+)}\to \bar Kωe^+ν_e$ and Determination of the Branching Fraction of $\bar K_1(1270)\to \bar K ω$

Using 20.3~fb$^{-1}$ of $e^+e^-$ annihilation data collected at a center-of-mass energy of 3.773~GeV with the BESIII detector, we report the first observation of the semileptonic decays $D^0\to K^-ωe^+ν_e$ and $D^+\to K_S^0ωe^+ν_e$ with significances of $8.0σ$ and $5.8σ$, respectively, including systematic uncertainties. Their decay branching fractions are measured to be ${\cal B}(D^0\to K^-ωe^+ν_e)=(9.3^{+2.1}_{-1.9}\pm 0.7)\times10^{-5}$ and ${\cal B}(D^+\to K_S^0ωe^+ν_e)=(6.6^{+2.0}_{-1.8}\pm 0.6)\times10^{-5}$. Combining with the latest measurements of $D^{0(+)}\to K^-π^+π^{-(0)} e^+ν_e$ and assuming $\bar{K}_1(1270)$ to be the sole mediating resonance in all processes, the branching ratios are determined to be $\frac{Γ(K_1(1270)^-\to K^-π^+π^-)}{Γ(K_1(1270)^-\to K^-ω)} = 3.4^{+0.8}_{-0.7} \pm 0.3$ and $\frac{Γ(\bar{K}_1(1270)^0\to K^-π^+π^0)}{Γ(\bar{K}_1(1270)^0\to \bar{K}^0ω)} = 9.6^{+3.0}_{-2.7} \pm 0.8$. The combined branching fraction is determined to be $\mathcal B(\bar{K}_1(1270)\to \bar{K}ω) = (7.5\pm 1.3 \pm 0.5)\%$, which is the most precise measurement from a collider experiment. The first uncertainties are statistical, and the second are systematic.

preprint2026arXiv

Guide, Think, Act: Interactive Embodied Reasoning in Vision-Language-Action Models

In this paper, we propose GTA-VLA(Guide, Think, Act), an interactive Vision-Language-Action (VLA) framework that enables spatially steerable embodied reasoning by allowing users to guide robot policies with explicit visual cues. Existing VLA models learn a direct "Sense-to-Act" mapping from multimodal observations to robot actions. While effective within the training distribution, such tightly coupled policies are brittle under out-of-domain (OOD) shifts and difficult to correct when failures occur. Although recent embodied Chain-of-Thought (CoT) approaches expose intermediate reasoning, they still lack a mechanism for incorporating human spatial guidance, limiting their ability to resolve visual ambiguities or recover from mistakes. To address this gap, our framework allows users to optionally guide the policy with spatial priors, such as affordance points, boxes, and traces, which the subsequent reasoning process can directly condition on. Based on these inputs, the model generates a unified spatial-visual Chain-of-Thought that integrates external guidance with internal task planning, aligning human visual intent with autonomous decision-making. For practical deployment, we further couple the reasoning module with a lightweight reactive action head for efficient action execution. Extensive experiments demonstrate the effectiveness of our approach. On the in-domain SimplerEnv WidowX benchmark, our framework achieves a state-of-the-art 81.2% success rate. Under OOD visual shifts and spatial ambiguities, a single visual interaction substantially improves task success over existing methods, highlighting the value of interactive reasoning for failure recovery in embodied control. Details of the project can be found here: https://signalispupupu.github.io/GTA-VLA_ProjPage/

preprint2026arXiv

LAGO: Language-Guided Adaptive Object-Region Focus for Zero-Shot Visual-Text Alignment

Zero-shot recognition aims to classify an image by selecting the most compatible label description from a set of candidate classes without any task-specific supervision. In fine-grained settings, however, the relevant evidence often lies in localized parts, attributes, or textures rather than in the full image, making whole-image alignment suboptimal. Recent localized visual-text alignment methods address this by comparing class descriptions with multiple image regions, but they typically rely on large sets of random or redundant crops, increasing inference cost and introducing many highly redundant or weakly relevant candidates. Moreover, introducing semantic guidance too early can create an error-amplifying feedback process in which inaccurate intermediate predictions bias later localization and reinforce subsequent mistakes; we refer to this failure mode as the prediction loop. We propose LAGO (LAnguage-Guided adaptive Object-region focus), a framework for efficient and robust zero-shot localized visual-text alignment. LAGO first performs class-agnostic object-centric candidate discovery to obtain a stable visual initialization, and then applies adaptive language-guided refinement with the strength of semantic guidance controlled by intermediate confidence. It further combines object-level, contextual, and full-image evidence through an effective object-context dual-channel aggregation strategy. Extensive experiments show that LAGO consistently achieves state-of-the-art performance on standard zero-shot benchmarks and challenging distribution-shift settings, while requiring substantially fewer candidate regions at inference time.

preprint2026arXiv

Learning from Prompt itself: the Hierarchical Attribution Prompt Optimization

Optimization is fundamental across numerous disciplines, typically following an iterative process of refining an initial solution to enhance performance. This principle is equally critical in prompt engineering, where designing effective prompts for large language models constitutes a complex optimization challenge. A structured optimization approach requires automated or semi-automated procedures to develop improved prompts, thereby reducing manual effort, improving performance, and yielding an interpretable process. However, current prompt optimization methods often induce prompt drift, where new prompts fix prior failures but impair performance on previously successful tasks. Additionally, generating prompts from scratch can compromise interpretability. To address these limitations, this study proposes the Hierarchical Attribution Prompt Optimization (HAPO) framework, which introduces three innovations: (1) a dynamic attribution mechanism targeting error patterns in training data and prompting history, (2) semantic-unit optimization for editing functional prompt segments, and (3) multimodal-friendly progression supporting both end-to-end LLM and LLM-MLLM workflows. Applied in contexts like single/multi-image QA (e.g., OCRV2) and complex task analysis (e.g., BBH), HAPO demonstrates enhanced optimization efficiency, outperforming comparable automated prompt optimization methods and establishing an extensible paradigm for scalable prompt engineering.

preprint2026arXiv

MAP: A Map-then-Act Paradigm for Long-Horizon Interactive Agent Reasoning

Current interactive LLM agents rely on goal-conditioned stepwise planning, where environmental understanding is acquired reactively during execution rather than established beforehand. This temporal inversion leads to Delayed Environmental Perception: agents must infer environmental constraints through trial-and-error, resulting in an Epistemic Bottleneck that traps them in inefficient failure cycles. Inspired by human affordance perception and cognitive map theory, we propose the Map-then-Act Paradigm (MAP), a plug-and-play framework that shifts environment understanding before execution. MAP consists of three stages: (1) Global Exploration, acquiring environment-general priors; (2) Task-Specific Mapping, constructing a structured cognitive map; and (3) Knowledge-Augmented Execution, solving tasks grounded on the map. Experiments show consistent gains across benchmarks and LLMs. On ARC-AGI-3, MAP enables frontier models to surpass near-zero baseline performance in 22 of 25 game environments. We further introduce MAP-2K, a dataset of map-then-act trajectories, and show that training on it outperforms expert execution traces, suggesting that understanding environments is more fundamental than imitation.

preprint2026arXiv

Measurements of the absolute branching fractions of the $Λ_{c}^{+}$ hadronic decays

Based on 4.5 fb$^{-1}$ of $e^+e^-$ collision data collected at center-of-mass energies between 4599.53 MeV and 4698.82 MeV with the BESIII detector, the absolute branching fractions of twelve $Λ_{c}^{+}$ hadronic decay modes are measured with a double-tag technique. A global least-square fit is implemented simultaneously among different decay modes at different energy points. This paper gives the most precise results on the branching fractions of different decay modes to date, with precision improved by a factor of 2 to 3. Among them, the branching fraction of $Λ_{c}^{+}\to pK^{-}π^+$ is determined to be $(6.61\pm0.11\pm0.12)\%$, where the first uncertainty is statistical and the second is systematic. In addition, the $e^+e^-\toΛ_c^+\barΛ_c^-$ Born cross sections and the effective form factors ($|G_{\rm eff}|$) at different energy points have been determined with the highest precision to date.

preprint2026arXiv

Measurements of the branching fractions of $χ_{cJ}\to 2K^+ 2K^- ω$ and $ϕK^+ K^- ω$ decays

Using a data sample of $(2712.4 \pm 14.3) \times 10^{6}$ $ψ(3686)$ events collected with the BESIII detector operating at the BEPCII collider, we report the first observation of the decays $χ_{cJ}\to 2K^+ 2K^- ω$ and $χ_{cJ}\to ϕK^{+}K^{-} ω$ ($J = 0,1,2$) via the radiative transitions $ψ(3686) \to γχ_{cJ}$. The branching fractions of these decays are measured for the first time, and the statistical significance for each signal exceeds $10σ$.

preprint2026arXiv

Measuring high-precision luminosity at the CEPC

Purpose: Luminosity measurement at the Circular Electron-Positron Collider (CEPC) is required to achieve 10^-4 precision when operating at the center-of-mass energy of the Z-pole. Approximately 10^12 Z-bosons will be collected to refine measurements of Standard Model processes. The design of the luminosity calorimeter (LumiCal) takes into account the geometry of the Machine-Detector-Interface (MDI) for the detection of Bhabha events. The detector simulation with GEANT predicts measurements of scattered electrons, positrons, and radiation photons. Results: The luminosity measurement derived from Bhabha event counting relies on the low-θ fiducial edge with a mean of better than 1 μRad. Both the beam monitoring on the interaction point (IP) and the LumiCal Si-wafer positions shall be monitored to a mean of better than 1 μm. The beam-pipe design is optimized with a low-mass window of less than 2 mm thick Be window for calibration of multiple scattering. With Si-layers capable of 5 μm resolution, the error on the mean of fiducial edges is measured to 1 μm. The detector displacement requires survey monitoring to sub-micron precision. Conclusion: The scattered electrons at IP are measured with the LumiCal Si-wafers and high granularity of LYSO bars. The accompanying photon with larger opening angles can be identified and studied for radiative Bhabha events. The NLO calculations for the Bhabha interaction are achieving 10^-4. With the LumiCal design of silicon detectors and LYSO calorimeters, the precision is pursued for IP and detector positions being monitored, to achieve the goal of 10^-4 precision on luminosity measurement.

preprint2026arXiv

MegaFlow: Large-Scale Distributed Orchestration System for the Agentic Era

The rapid development of interactive and autonomous AI systems signals our entry into the agentic era. Training and evaluating agents on complex agentic tasks such as software engineering and computer use requires not only efficient model computation but also sophisticated infrastructure capable of coordinating vast agent-environment interactions. However, no open-source infrastructure can effectively support large-scale training and evaluation on such complex agentic tasks. To address this challenge, we present MegaFlow, a large-scale distributed orchestration system that enables efficient scheduling, resource allocation, and fine-grained task management for agent-environment workloads. MegaFlow abstracts agent training infrastructure into three independent services (Model Service, Agent Service, and Environment Service) that interact through unified interfaces, enabling independent scaling and flexible resource allocation across diverse agent-environment configurations. In our agent training deployments, MegaFlow successfully orchestrates tens of thousands of concurrent agent tasks while maintaining high system stability and achieving efficient resource utilization. By enabling such large-scale agent training, MegaFlow addresses a critical infrastructure gap in the emerging agentic AI landscape.

preprint2026arXiv

Modulating near-field radiative energy and momentum transfer via rotating Weyl semimetals

We study near-field radiative transfer of energy, angular momentum, and linear momentum between a nanoparticle and a plate consisting of magnetic Weyl semimetals, and demonstrate that these can be efficiently tuned by a relative angle between the Weyl node separations. This tunability originates from the coupling between the particle-induced rotational Poynting vector and the nonreciprocal surface plasmon polaritons supported by the plate. Remarkably, we uncover a counterintuitive regime in which both energy and angular momentum transfer are maximized when the Weyl node separations are antiparallel rather than parallel. This arises from optimal mode matching between the rotation direction of the particle's circular heat flux and the propagation direction of the surface plasmon polaritons in the antiparallel configuration.

preprint2026arXiv

Numerical analysis of spatiotemporal high-index saddle dynamics for finding multiple solutions of semilinear elliptic problems

This paper presents a rigorous numerical framework for computing multiple solutions of semilinear elliptic problems by spatiotemporal high-index saddle dynamics (HiSD), which extends the traditional HiSD to the continuous-in-space setting, explicitly incorporating spatial differential operators. To enforce the Stiefel manifold constraint without introducing the analytical complications of retraction-based updates, we design a fully discrete retraction-free orthonormality-preserving scheme for spatiotemporal HiSD. This scheme exhibits favorable structural properties that substantially reduce the difficulties arising from coupling and gradient nonlinearities in spatiotemporal HiSD. Exploiting these properties, we establish gradient stability and error estimates, which consequently ensure the preservation of the Morse index for the computed saddle points. The framework is further extended to the semilinear advection-reaction-diffusion equation. Numerical experiments demonstrate the efficiency of the proposed method in finding multiple solutions and constructing the solution landscape of semilinear elliptic problems. To the best of our knowledge, this work presents the first rigorous full space--time accuracy analysis of the HiSD system. It reveals intrinsic connections between saddle-search algorithms and numerical methods for PDEs, enhancing their mutual compatibility for a broad range of problems.

preprint2026arXiv

Observation of Polarization and Determination of Electric and Magnetic Moments of $Ξ(1530)^0$ in $ψ(3686)\toΞ(1530)^0\barΞ(1530)^0$

Using the data sample of $2.7\times10^9$ $ψ(3686)$ events collected with the BESIII detector at the BEPCII collider, we present an observation of the $Ξ(1530)^0$ polarization in the decay $ψ(3686)\toΞ(1530)^0\barΞ(1530)^0$ with a significance larger than $20σ$ compared with all other tested hypotheses. The helicity amplitudes for the process $ψ(3686)\toΞ(1530)^0\barΞ(1530)^0$ and the moduli of form factors including electric charge, magnetic dipole, electric quadrupole, and magnetic octupole are measured for the first time by performing an angular distribution analysis. Additionally, the polarization correlations between $Ξ(1530)^0$ and $\barΞ(1530)^0$ are measured.

preprint2026arXiv

OVSeg3R: Learn Open-vocabulary Instance Segmentation from 2D via 3D Reconstruction

In this paper, we propose a training scheme called OVSeg3R to learn open-vocabulary 3D instance segmentation from well-studied 2D perception models with the aid of 3D reconstruction. OVSeg3R directly adopts reconstructed scenes from 2D videos as input, avoiding costly manual adjustment while aligning input with real-world applications. By exploiting the 2D to 3D correspondences provided by 3D reconstruction models, OVSeg3R projects each view's 2D instance mask predictions, obtained from an open-vocabulary 2D model, onto 3D to generate annotations for the view's corresponding sub-scene. To avoid incorrectly introduced false positives as supervision due to partial annotations from 2D to 3D, we propose a View-wise Instance Partition algorithm, which partitions predictions to their respective views for supervision, stabilizing the training process. Furthermore, since 3D reconstruction models tend to over-smooth geometric details, clustering reconstructed points into representative super-points based solely on geometry, as commonly done in mainstream 3D segmentation methods, may overlook geometrically non-salient objects. We therefore introduce 2D Instance Boundary-aware Superpoint, which leverages 2D masks to constrain the superpoint clustering, preventing superpoints from violating instance boundaries. With these designs, OVSeg3R not only extends a state-of-the-art closed-vocabulary 3D instance segmentation model to open-vocabulary, but also substantially narrows the performance gap between tail and head classes, ultimately leading to an overall improvement of +2.3 mAP on the ScanNet200 benchmark. Furthermore, under the standard open-vocabulary setting, OVSeg3R surpasses previous methods by about +7.1 mAP on the novel classes, further validating its effectiveness.

preprint2026arXiv

PlotCraft: Pushing the Limits of LLMs for Complex and Interactive Data Visualization

Recent Large Language Models (LLMs) have demonstrated remarkable proficiency in code generation. However, their ability to create complex visualizations for scaled and structured data remains largely unevaluated and underdeveloped. To address this gap, we introduce PlotCraft, a new benchmark featuring 1k challenging visualization tasks that cover a wide range of topics, such as finance, scientific research, and sociology. The benchmark is structured around seven high-level visualization tasks and encompasses 48 distinct chart types. Crucially, it is the first to systematically evaluate both single-turn generation and multi-turn refinement across a diverse spectrum of task complexities. Our comprehensive evaluation of 23 leading LLMs on PlotCraft reveals obvious performance deficiencies in handling sophisticated visualization tasks. To bridge this performance gap, we develope SynthVis-30K, a large-scale, high-quality dataset of complex visualization code synthesized via a collaborative agent framework. Building upon this dataset, we develope PlotCraftor, a novel code generation model that achieves strong capabilities in complex data visualization with a remarkably small size. Across VisEval, PandasPlotBench, and our proposed PlotCraft, PlotCraftor shows performance comparable to that of leading proprietary approaches. Especially, on hard task, Our model achieves over 50% performance improvement. We will release the benchmark, dataset, and code at https://github.com/Speakn0w/PlotCraft-Benchmark.

preprint2026arXiv

Programmable calculus operations in electromagnetic space using space-time-coding metasurface

With the rapid advancement of metasurfaces and the increasing demand for programmable metasurfaces to simplify information systems, wave-based computation using metasurfaces has emerged as an attractive research topic. To facilitate the mathematical operations in electromagnetic (EM) space, here we propose a space-time coding metasurface (STCM) system capable of directly performing calculus operations on the spatial energy distributions of EM waves. By exploiting harmonic characteristics induced by time-varying coding, the responses of meta-atoms at specific harmonics can be flexibly controlled, which enables the metasurface system to address more complex tasks. Owing to its programmability, the STCM can dynamically switch functions in real time to accommodate different calculus tasks. To fully leverage the capability of STCM, we not only present the space-time coding sequences for differentiation and integration of EM waves, but also develop and numerically simulate the space-time coding sequences that can independently and simultaneously implement different calculus operations on the same incident EM waves. To experimentally validate the feasibility of the EM calculus operations, proof-of-concept experiments are conducted using a programmable 2-bit STCM. Good agreements among the theory, numerical simulations, and experiments confirm the feasibility of performing calculus operations in the EM space and demonstrate the broad application prospects of STCM in EM wave manipulations, wireless communications, and signal processing.

preprint2026arXiv

Pulsar Polarization Array Limits on Ultralight Axion-like Dark Matter

We conduct the first-ever Pulsar Polarization Array (PPA) analysis to detect the ultralight Axion-Like Dark Matter (ALDM) using the polarization data of 22 millisecond pulsars from the third data release of Parkes Pulsar Timing Array. As one of the major dark matter candidates, the ultralight ALDM exhibits a pronounced wave nature on astronomical scales and offers a promising solution to small-scale structure issues within local galaxies. While the linearly polarized pulsar light travels through the ALDM galactic halo, its position angle (PA) can be subject to an oscillation induced by the ALDM Chern-Simons coupling with electromagnetic field. The PPA is thus especially suited for detecting the ultralight ALDM by correlating polarization data across the arrayed pulsars. To accomplish this task, we develop an advanced Bayesian analysis framework that allows us to construct pulsar PA residual time series, model noise contributions properly and search for pulsar cross-correlations. We find that for an ALDM density of $ρ_0=0.4\,\textrm{GeV}/\textrm{cm}^3$, the Parkes PPA offers the best global limits on the ALDM Chern-Simons coupling, namely $\lesssim 10^{-13.5}-10^{-12.2}~{\rm GeV}^{-1}$, for the mass range of $10^{-22} - 10^{-21}~{\rm eV}$. The crucial role of pulsar cross-correlation in recognizing the nature of the derived limits is also highlighted.

preprint2026arXiv

Radiation Resistance of Ge-doped Multi-Mode Fiber for Optical Links in Collider Experiments

The applications of optical links in collider experiments provide the advantage of high-speed data transmission with low mass fibers over distances of a few hundred meters. Ge-doped multi-mode fibers are evaluated for radiation tolerance in ionizing doses of Co-60 gamma rays. The Radiation-Induced Attenuation (RIA) varies significantly depending on doping substances and fabrication technologies. A type of telecom-grade fiber has demonstrated an RIA of 0.05 dB/m under a total ionizing dose of 300 kGy(SiO2). The dependence on dose rate is compared in the range between 5 Gy/hr and 1.4 kGy/hr, and the annealing recovery is observed after the Co-60 source is shielded. The temperature dependence is investigated across a range of -15 oC to room temperature. At cold temperatures, stagnant annealing leads to a substantially higher RIA during irradiation. The recovery of radiation-induced defects is typically within a few hours, resulting in similar RIA levels regardless of the dose rate and temperature during exposure. Ge-doped fibers of chosen fabrication methods are capable of enduring high ionizing doses for use in high-energy physics experiments.

preprint2026arXiv

SaddleScape V1.0: A Python Package for Constructing Solution Landscapes via High-index Saddle Dynamics

We present SaddleScape V1.0, a Python software package designed for the exploration and construction of solution landscapes in complex systems. The package implements the High-index Saddle Dynamics (HiSD) framework and its variants, including the Generalized HiSD for non-gradient systems and the Accelerated HiSD. SaddleScape V1.0 enables the systematic identification of critical points, including both local minima and high-index saddle points, by dynamically updating both the state estimate and an associated subspace characterizing the saddle's local manifold. It supports both gradient systems, defined by energy functions/functionals, and general non-gradient autonomous dynamical systems. Key features include automatic differentiation for symbolic inputs, numerical approximation techniques for Hessian-vector products, diverse eigenvalue solvers, and algorithms for constructing solution landscapes. The software offers a user-friendly interface with flexible parameter configuration, tools for trajectory and landscape visualization, and data export capabilities. By providing an efficient and accessible implementation of advanced saddle dynamics, SaddleScape V1.0 facilitates the construction of solution landscapes, empowering researchers in various scientific disciplines to gain deeper insights into the hierarchical structure of complex systems. The source code is available at the repository https://github.com/HiSDpackage/saddlescape. The package's introductory website is available at https://hisdpackage.github.io/saddlescape.

preprint2026arXiv

Search for a dark baryon in the $Ξ^-\rightarrowπ^-+{\rm invisible}$ decay

A search for a dark baryon is performed for the first time in the two-body decay $Ξ^-\rightarrowπ^-+{\rm invisible}$ using $(10.087\pm0.044)\times10^{9}$ $J/ψ$ events collected at a center-of-mass energy of $\sqrt{s}=3.097\,\mbox{GeV}$ with the BESIII detector at the BEPCII collider. No significant signal is observed, and the 90% (95%) confidence level upper limits on the branching fraction $B(Ξ^-\rightarrowπ^-+{\rm invisible})$ are determined to be $4.2\times10^{-5}$ ($5.2\times10^{-5}$), $6.9\times10^{-5}$ ($8.4\times10^{-5}$), $6.5\times10^{-4}$ ($7.6\times10^{-4}$), $1.1\times10^{-4}$ ($1.3\times10^{-4}$) and $4.5\times10^{-5}$ ($5.5\times10^{-5}$), under the dark baryon mass hypotheses of 1.07$\,\mbox{GeV}/c^2$, 1.10$\,\mbox{GeV}/c^2$, $m_Λ$ (1.116$\,\mbox{GeV}/c^2$), 1.13$\,\mbox{GeV}/c^2$, and 1.16$\,\mbox{GeV}/c^2$, respectively. The constraints obtained on the Wilson coefficients $C_{u s, s}^L$ and $C_{u s, s}^R$ are more stringent than the previous limits derived from the LHC searches for the colored mediators.

preprint2026arXiv

UM3: Unsupervised Map to Map Matching

Map-to-map matching is a critical task for aligning spatial data across heterogeneous sources, yet it remains challenging due to the lack of ground truth correspondences, sparse node features, and scalability demands. In this paper, we propose an unsupervised graph-based framework that addresses these challenges through three key innovations. First, our method is an unsupervised learning approach that requires no training data, which is crucial for large-scale map data where obtaining labeled training samples is challenging. Second, we introduce pseudo coordinates that capture the relative spatial layout of nodes within each map, which enhances feature discriminability and enables scale-invariant learning. Third, we design an mechanism to adaptively balance feature and geometric similarity, as well as a geometric-consistent loss function, ensuring robustness to noisy or incomplete coordinate data. At the implementation level, to handle large-scale maps, we develop a tile-based post-processing pipeline with overlapping regions and majority voting, which enables parallel processing while preserving boundary coherence. Experiments on real-world datasets demonstrate that our method achieves state-of-the-art accuracy in matching tasks, surpassing existing methods by a large margin, particularly in high-noise and large-scale scenarios. Our framework provides a scalable and practical solution for map alignment, offering a robust and efficient alternative to traditional approaches.

preprint2026arXiv

Uncertainty-Aware Exploratory Direct Preference Optimization for Multimodal Large Language Models

Direct Preference Optimization (DPO) has proven to be an effective solution for mitigating hallucination in Multimodal Large Language Models (MLLMs) by learning from preference pairs. One of its key challenges lies in how to transfer the sequence-level preference into fine-grained supervision on visual fidelity. To safeguard vision-related tokens that are prone to hallucination, existing methods typically allocate training emphasis according to the model's self-assessed visual sensitivity signals. However, such sensitivity, estimated by a model still under training, introduces self-referential bias: reinforcing already well-learned visual cues while neglecting hard-to-perceive but critical details, thereby limiting deeper alignment. In this work, we propose an Uncertainty-aware Exploratory Direct Preference Optimization (UE-DPO) method for MLLMs, which enables the model to uncover its cognitive deficiencies and actively explore for self-correction, guided by token-level epistemic uncertainty. Specifically, we first quantify the uncertainty from the model's failure to ground token predictions in the given image. Then, based on an uncertainty-aware exploration intensity, we encourage more learning pressure on visually deficient tokens in preferred samples, and alleviate the over-penalization of beneficial knowledge in dispreferred samples. Further, we provide a theoretical justification for our method, and extensive experiments demonstrate its effectiveness and robustness.

preprint2026arXiv

UniTriGen: Unified Triplet Generation of Aligned Visible-Infrared-Label for Few-Shot RGB-T Semantic Segmentation

RGB-T semantic segmentation requires strictly aligned VIS-IR-Label triplets; however, such aligned triplet data are often scarce in real-world scenarios. Existing generative augmentation methods usually adopt cascaded generation paradigms, decomposing joint triplet generation into local conditional processes. As a result, consistency among VIS, IR, and Label in spatial structure, semantic content, and cross-modal details cannot be reliably maintained. To address this issue, we propose UniTriGen, a unified triplet generation framework that directly generates spatially aligned, semantically consistent, and modality complementary VIS-IR-Label triplets under the guidance of text prompts. UniTriGen first introduces a unified triplet generation mechanism, where VIS, IR, and Label are jointly encoded into a shared latent space and modeled with a diffusion process to enforce global cross-modal consistency. Lightweight modality-specific residual adapters are further integrated into this mechanism to accommodate modality-specific imaging characteristics and output formats. To mitigate generation bias caused by imbalanced scene and class distributions in limited paired triplets, UniTriGen also employs a scene-balanced and class-aware few-shot sampling strategy, which induces a more balanced sampling distribution and enhances the scene and class diversity of generated triplets. Experiments show that UniTriGen generates high-quality aligned triplets from limited real paired data, thereby achieving consistent performance improvements across various RGB-T semantic segmentation models.

preprint2026arXiv

Weighted Reverse Convolution for Feature Upsampling

Pre-trained vision foundation models (VFMs) provide strong semantic representations, yet their patch-level features are inherently coarse, limiting their effectiveness on tasks requiring fine-grained localization, dense prediction, and point-wise correspondence. In this work, we revisit feature upsampling for VFMs from the perspective of \textbf{\textit{inverse problem}} and propose Weighted Reverse Convolution (WRC), a spatially adaptive inverse operator for densifying high-level visual descriptors. Specifically, we formulate feature upsampling as a weighted Tikhonov-regularized least-squares problem, where spatially varying weights modulate both data fidelity and prior strength at each spatial location. This allows WRC to adapt the reconstruction to spatially varying feature characteristics, thereby preserving critical structures while mitigating over-smoothing. Moreover, WRC retains an efficient, fully differentiable closed-form FFT solution, making it a practical drop-in upsampling operator. Integrated into a lightweight self-supervised densification framework, WRC consistently improves dense feature quality across various downstream benchmarks, including segmentation, depth estimation, video object segmentation, object discovery, and keypoint correspondence, while maintaining high computational efficiency.

preprint2026arXiv

Which Face and Whose Identity? Solving the Dual Challenge of Deepfake Proactive Forensics in Multi-Face Scenarios

Unlike single-face forgeries, deepfakes in complex multi-person interaction scenarios (such as group photos and multi-person meetings) more closely reflect real-world threats. Although existing proactive forensics solutions demonstrate good performance, they heavily rely on a "single-face" setting, making it difficult to effectively address the problems of deepfake localization and source tracing in complex multi-person environments. To address this challenge, we propose the Deep Attributable Watermarking Framework (DAWF). This framework adopts a novel multi-face encoder-decoder architecture that bypasses the cumbersome offline pre-processing steps of traditional forensics, facilitating efficient in-network parallel watermark embedding and cross-face collaborative processing. Crucially, we propose a selective regional supervision loss. This innovative mechanism guides the decoder to focus exclusively on the facial regions tampered with by deepfakes. Leveraging this mechanism alongside the embedded identity payloads, DAWF realizes the "which + who" goal, answering the dual questions of which facial region was forged and who was forged. Extensive experiments on challenging multi-face datasets show that DAWF achieves excellent deepfake localization and traceability in complex multi-person scenes.

preprint2026arXiv

Xiaomi EV World Model: A Joint World Model Integrating Reconstruction and Generation for Autonomous Driving

This report presents a unified technical system addressing the two core capabilities of world models for autonomous driving: world representation and world generation. For world representation, we propose WorldRec, a feed-forward reconstruction architecture driven by sparse scene queries. WorldRec initializes structured queries in 3D space, leveraging them to aggregate cross-view, cross-temporal features, thereby naturally enforcing spatial consistency across frames and yielding compact yet high-fidelity 3D Gaussian scene representations. For world generation, we propose WorldGen, a two-stage training framework of bidirectional pretraining followed by causal fine-tuning through three progressive stages (Teacher Forcing, ODE distillation, and DMD), enabling high-quality online causal video generation in as few as 4 denoising steps. Building on both modules, we further introduce the JWM, which deeply integrates WorldRec and WorldGen to achieve synergistic gains in generation stability, cross-frame consistency, and visual fidelity, providing a solid foundation for closed-loop simulation, data synthesis, and end-to-end training in autonomous driving.

preprint2025arXiv

Measurement of branching fractions of $Λ_{c}^{+}$ decays to $Σ^{+} η$ and $Σ^{+} η'$

By analyzing $e^+e^-$ collision data taken at center-of-mass energies $\sqrt{s}$ between 4.600 and 4.699 GeV with the BESIII detector at the BEPCII collider, corresponding to an integrated luminosity of $\rm 4.5~fb^{-1}$, we study the hadronic decays $Λ_{c}^{+} \rightarrow Σ^{+} η$ and $Λ_{c}^{+} \rightarrow Σ^{+} η^{\prime}$ using the single-tag method. The branching fraction ratio of $Λ_{c}^+ \rightarrow Σ^+ η$ relative to $Λ_{c}^+ \rightarrow Σ^+ π^0$ is determined to be $0.305 \pm 0.046_{\rm stat.} \pm 0.007_{\rm syst.}$, and that of $Λ_{c}^+ \rightarrow Σ^+ η'$ relative to $Λ_{c}^+ \rightarrow Σ^+ ω$ is $0.336 \pm 0.094_{\rm stat.} \pm 0.037_{\rm syst.}$. The ratio of $\frac{\mathcal{B}\left(Λ_{c}^{+} \rightarrow Σ^{+} η'\right)}{\mathcal{B}\left(Λ_{c}^{+} \rightarrow Σ^{+} η\right)} $ is determined to be $1.73 \pm 0.22_{\rm stat.} \pm 0.16_{\rm syst.}$. These results enrich our knowledge of charmed baryon decays.

preprint2025arXiv

Ultra-Wideband Polarimetry of the April 2021 Profile Change Event in PSR J1713+0747

The millisecond pulsar PSR J1713+0747 is a high-priority target for pulsar timing array experiments due to its long-term timing stability, and bright, narrow pulse profile. In April 2021, PSR~J1713$+$0747 underwent a significant profile change event, observed by several telescopes worldwide. Using the broad-bandwidth and polarimetric fidelity of the Ultra-Wideband Low-frequency receiver on Murriyang, CSIRO's Parkes radio telescope, we investigated the long-term spectro-polarimetric behaviour of this profile change in detail. We highlight the broad-bandwidth nature of the event, which exhibits frequency dependence that is inconsistent with cold-plasma propagation effects. We also find that spectral and temporal variations are stronger in one of the orthogonal polarisation modes than the other, and observe mild variations ($\sim 3$ - $5\,σ$ significance) in circular polarisation above 1400 MHz following the event. However, the linear polarisation position angle remained remarkably stable in the profile leading edge throughout the event. With over three years of data post-event, we find that the profile has not yet recovered back to its original state, indicating a long-term asymptotic recovery, or a potential reconfiguration of the pulsar's magnetic field. These findings favour a magnetospheric origin of the profile change event over a line-of-sight propagation effect in the interstellar medium.

preprint2024arXiv

Investigation of the $ΔI = 1/2$ rule and test of CP violation through the measurement of decay asymmetry parameters in $Ξ^-$ decays

Using $(10087\pm44)\times 10^{6}$ $J/ψ$ events collected with the BESIII detector, numerous $Ξ^-$ and $Λ$ decay asymmetry parameters are simultaneously determined from the process $J/ψ\to Ξ^- \barΞ^+ \to Λ(pπ^-) π^- \barΛ(\bar{n} π^0) π^+$ and its charge-conjugate channel. The precisions of $α_0$ for $Λ\to nπ^0$ and $\barα_0$ for $\barΛ \to \bar{n}π^0$ compared to world averages are improved by factors of 4 and 1.7, respectively. The ratio of decay asymmetry parameters of $Λ\to nπ^0$ to that of $Λ\to pπ^-$, $\langle α_0 \rangle/ \langle α_{Λ-} \rangle $, is determined to be $ 0.873 \pm 0.012^{+0.011}_{-0.010}$, where the first and the second uncertainties are statistical and systematic, respectively. The ratio is smaller than unity more than $5σ$, which signifies the existence of the $ΔI = 3/2$ transition in $Λ$ for the first time. Beside, we test for CP violation in $Ξ^- \to Λπ^-$ and in $Λ\to n π^{0}$ with the best precision to date.

preprint2024arXiv

Robust Beamforming Design for Intelligent Reflecting Surface Aided Cognitive Radio Systems with Imperfect Cascaded CSI

In this paper, intelligent reflecting surface (IRS) is introduced to enhance the network performance of cognitive radio (CR) systems. Specifically, we investigate robust beamforming design based on both bounded channel state information (CSI) error model and statistical CSI error model for primary user (PU)-related channels in IRS-aided CR systems. We jointly optimize the transmit precoding (TPC) at the secondary user (SU) transmitter (ST) and phase shifts at the IRS to minimize the ST' s total transmit power subject to the quality of service of SUs, the limited interference imposed on the PU and unit-modulus of the reflective beamforming. The successive convex approximation (SCA) method, Schur's complement, General sign-definiteness principle, inverse Chi-square distribution and penalty convex-concave procedure are invoked for dealing with these intricate constraints. The non-convex optimization problems are transformed into several convex subproblems and efficient algorithms are proposed. Simulation results verify the efficiency of the proposed algorithms and reveal the impacts of CSI uncertainties on ST's minimum transmit power and feasibility rate of the optimization problems. Simulation results also show that the number of transmit antennas at the ST and the number of phase shifts at the IRS should be carefully chosen to balance the channel realization feasibility rate and the total transmit power.

preprint2024arXiv

Testing Complex Singlet Scalar Cosmology at the Large Hadron Collider

The Standard Model extended with a complex singlet scalar (cxSM) can admit a strong first order electroweak phase transition (SFOEWPT) as needed for electroweak baryogenesis and provide a dark matter (DM) candidate. The presence of both a DM candidate and a singlet-like scalar that mixes with the Standard Model Higgs boson leads to the possibility of a $b\bar{b}+\text{MET}$ final state in $pp$ collisions. Focusing on this channel, we analyze the prospective reach at the Large Hadron Collider (LHC) for a heavy singlet-like scalar in regions of cxSM parameter space compatible with a SFOEWT and DM phenomenology. We identify this parameter space while implementing current constraints from electroweak precision observable and Higgs boson property measurements as well as those implied by LHC heavy resonance searches.

preprint2023arXiv

Human-Timescale Adaptation in an Open-Ended Task Space

Foundation models have shown impressive adaptation and scalability in supervised and self-supervised learning problems, but so far these successes have not fully translated to reinforcement learning (RL). In this work, we demonstrate that training an RL agent at scale leads to a general in-context learning algorithm that can adapt to open-ended novel embodied 3D problems as quickly as humans. In a vast space of held-out environment dynamics, our adaptive agent (AdA) displays on-the-fly hypothesis-driven exploration, efficient exploitation of acquired knowledge, and can successfully be prompted with first-person demonstrations. Adaptation emerges from three ingredients: (1) meta-reinforcement learning across a vast, smooth and diverse task distribution, (2) a policy parameterised as a large-scale attention-based memory architecture, and (3) an effective automated curriculum that prioritises tasks at the frontier of an agent's capabilities. We demonstrate characteristic scaling laws with respect to network size, memory length, and richness of the training task distribution. We believe our results lay the foundation for increasingly general and adaptive RL agents that perform well across ever-larger open-ended domains.

preprint2023arXiv

Search for hidden-charm tetraquark with strangeness in $e^{+}e^{-}\rightarrow K^+ D_{s}^{-} D^{0}+c.c.$

We report a search for a heavier partner of the recently observed $Z_{cs}(3985)^{-}$ state, denoted as $Z_{cs}^{\prime -}$, in the process $e^{+} e^{-}\rightarrow K^{+}D_{s}^{*-}D^{* 0}+c.c.$, based on $e^+e^-$ collision data collected at the center-of-mass energies of $\sqrt{s}=4.661$, 4.682 and 4.699 GeV with the BESIII detector. The $Z_{cs}^{\prime -}$ is of interest as it is expected to be a candidate for a hidden-charm and open-strange tetraquark. A partial-reconstruction technique is used to isolate $K^+$ recoil-mass spectra, which are probed for a potential contribution from $Z_{cs}^{\prime -}\to D_{s}^{*-}D^{* 0}$ ($c.c.$). We find an excess of $Z_{cs}^{\prime -}\rightarrow D_{s}^{*-}D^{*0}$ ($c.c.$) candidates with a significance of $2.1σ$, after considering systematic uncertainties, at a mass of $(4123.5\pm0.7_\mathrm{stat.}\pm4.7_\mathrm{syst.})\ \mathrm{MeV}/c^{2}$. As the data set is limited in size, the upper limits are evaluated at the 90\% confidence level on the product of the Born cross sections ($σ^{\mathrm{Born}}$) and the branching fraction ($\mathcal{B}$) of $Z_{cs}^{\prime-}\rightarrow D_{s}^{*-}D^{* 0}$, under different assumptions of the $Z_{cs}^{\prime -}$ mass from 4.120 to 4.140 MeV and of the width from 10 to 50 MeV at the three center-of-mass energies. The upper limits of $σ^{\rm Born}\cdot\mathcal{B}$ are found to be at the level of $\mathcal{O}(1)$ pb at each energy. Larger data samples are needed to confirm the $Z_{cs}^{\prime -}$ state and clarify its nature in the coming years.

preprint2022arXiv

A Differentiable Two-stage Alignment Scheme for Burst Image Reconstruction with Large Shift

Denoising and demosaicking are two essential steps to reconstruct a clean full-color image from the raw data. Recently, joint denoising and demosaicking (JDD) for burst images, namely JDD-B, has attracted much attention by using multiple raw images captured in a short time to reconstruct a single high-quality image. One key challenge of JDD-B lies in the robust alignment of image frames. State-of-the-art alignment methods in feature domain cannot effectively utilize the temporal information of burst images, where large shifts commonly exist due to camera and object motion. In addition, the higher resolution (e.g., 4K) of modern imaging devices results in larger displacement between frames. To address these challenges, we design a differentiable two-stage alignment scheme sequentially in patch and pixel level for effective JDD-B. The input burst images are firstly aligned in the patch level by using a differentiable progressive block matching method, which can estimate the offset between distant frames with small computational cost. Then we perform implicit pixel-wise alignment in full-resolution feature domain to refine the alignment results. The two stages are jointly trained in an end-to-end manner. Extensive experiments demonstrate the significant improvement of our method over existing JDD-B methods. Codes are available at https://github.com/GuoShi28/2StageAlign.

preprint2022arXiv

A Dual Weighting Label Assignment Scheme for Object Detection

Label assignment (LA), which aims to assign each training sample a positive (pos) and a negative (neg) loss weight, plays an important role in object detection. Existing LA methods mostly focus on the design of pos weighting function, while the neg weight is directly derived from the pos weight. Such a mechanism limits the learning capacity of detectors. In this paper, we explore a new weighting paradigm, termed dual weighting (DW), to specify pos and neg weights separately. We first identify the key influential factors of pos/neg weights by analyzing the evaluation metrics in object detection, and then design the pos and neg weighting functions based on them. Specifically, the pos weight of a sample is determined by the consistency degree between its classification and localization scores, while the neg weight is decomposed into two terms: the probability that it is a neg sample and its importance conditioned on being a neg sample. Such a weighting strategy offers greater flexibility to distinguish between important and less important samples, resulting in a more effective object detector. Equipped with the proposed DW method, a single FCOS-ResNet-50 detector can reach 41.5% mAP on COCO under 1x schedule, outperforming other existing LA methods. It consistently improves the baselines on COCO by a large margin under various backbones without bells and whistles. Code is available at https://github.com/strongwolf/DW.

preprint2022arXiv

A model-free shrinking-dimer saddle dynamics for finding saddle point and solution landscape

We propose a model-free shrinking-dimer saddle dynamics for finding any-index saddle points and constructing the solution landscapes, in which the force in the standard saddle dynamics is replaced by a surrogate model trained by the Gassian process learning. By this means, the exact form of the model is no longer necessary such that the saddle dynamics could be implemented based only on some observations of the force. This data-driven approach not only avoids the modeling procedure that could be difficult or inaccurate, but also significantly reduces the number of queries of the force that may be expensive or time-consuming. We accordingly develop a sequential learning saddle dynamics algorithm to perform a sequence of local saddle dynamics, in which the queries of the training samples and the update or retraining of the surrogate force are performed online and around the latent trajectory in order to improve the accuracy of the surrogate model and the value of each sampling. Numerical experiments are performed to demonstrate the effectiveness and efficiency of the proposed algorithm.

preprint2022arXiv

A Route Network Planning Method for Urban Air Delivery

High-tech giants and start-ups are investing in drone technologies to provide urban air delivery service, which is expected to solve the last-mile problem and mitigate road traffic congestion. However, air delivery service will not scale up without proper traffic management for drones in dense urban environment. Currently, a range of Concepts of Operations (ConOps) for unmanned aircraft system traffic management (UTM) are being proposed and evaluated by researchers, operators, and regulators. Among these, the tube-based (or corridor-based) ConOps has emerged in operations in some regions of the world for drone deliveries and is expected to continue serving certain scenarios that with dense and complex airspace and requires centralized control in the future. Towards the tube-based ConOps, we develop a route network planning method to design routes (tubes) in a complex urban environment in this paper. In this method, we propose a priority structure to decouple the network planning problem, which is NP-hard, into single-path planning problems. We also introduce a novel space cost function to enable the design of dense and aligned routes in a network. The proposed method is tested on various scenarios and compared with other state-of-the-art methods. Results show that our method can generate near-optimal route networks with significant computational time-savings.

preprint2022arXiv

A Survey on Leveraging Pre-trained Generative Adversarial Networks for Image Editing and Restoration

Generative adversarial networks (GANs) have drawn enormous attention due to the simple yet effective training mechanism and superior image generation quality. With the ability to generate photo-realistic high-resolution (e.g., $1024\times1024$) images, recent GAN models have greatly narrowed the gaps between the generated images and the real ones. Therefore, many recent works show emerging interest to take advantage of pre-trained GAN models by exploiting the well-disentangled latent space and the learned GAN priors. In this paper, we briefly review recent progress on leveraging pre-trained large-scale GAN models from three aspects, i.e., 1) the training of large-scale generative adversarial networks, 2) exploring and understanding the pre-trained GAN models, and 3) leveraging these models for subsequent tasks like image restoration and editing. More information about relevant methods and repositories can be found at https://github.com/csmliu/pretrained-GANs.

preprint2022arXiv

A Text Attention Network for Spatial Deformation Robust Scene Text Image Super-resolution

Scene text image super-resolution aims to increase the resolution and readability of the text in low-resolution images. Though significant improvement has been achieved by deep convolutional neural networks (CNNs), it remains difficult to reconstruct high-resolution images for spatially deformed texts, especially rotated and curve-shaped ones. This is because the current CNN-based methods adopt locality-based operations, which are not effective to deal with the variation caused by deformations. In this paper, we propose a CNN based Text ATTention network (TATT) to address this problem. The semantics of the text are firstly extracted by a text recognition module as text prior information. Then we design a novel transformer-based module, which leverages global attention mechanism, to exert the semantic guidance of text prior to the text reconstruction process. In addition, we propose a text structure consistency loss to refine the visual appearance by imposing structural consistency on the reconstructions of regular and deformed texts. Experiments on the benchmark TextZoom dataset show that the proposed TATT not only achieves state-of-the-art performance in terms of PSNR/SSIM metrics, but also significantly improves the recognition accuracy in the downstream text recognition task, particularly for text instances with multi-orientation and curved shapes. Code is available at https://github.com/mjq11302010044/TATT.

preprint2022arXiv

A theorem on meromorphic descent and the specialization of the pro-étale fundamental group

Given a Noetherian formal scheme $\hat X$ over ${\rm Spf}(R)$, where $R$ is a complete DVR, we first prove a theorem of meromorphic descent along a possibly infinite cover of $\hat{X}$. Using this we construct a specialization functor from the category of continuous representations of the pro-étale fundamental group of the special fiber to the category of $F$-divided sheaves on the generic fiber. This specialization functor partially recovers the specialization functor of the étale fundamental groups. We also express the pro-étale fundamental group of a connected scheme $X$ of finite type over a field as coproducts and quotients of the free group and the étale fundamental groups of the normalizations of the irreducible components of $X$ and those of its singular loci.

preprint2022arXiv

Adaptive Multigrid Strategy for Geometry Optimization of Large-Scale Three Dimensional Molecular Mechanics

In this paper, we present an efficient adaptive multigrid strategy for the geometry optimization of large-scale three dimensional molecular mechanics. The resulting method can achieve significantly reduced complexity by exploiting the intrinsic low-rank property of the material configurations and by combining the state-of-the-art adaptive techniques with the hierarchical structure of multigrid algorithms. To be more precise, we develop a oneway multigrid method with adaptive atomistic/continuum (a/c) coupling, e.g., blended ghost force correction (BGFC) approximations with gradient-based a posteriori error estimators on the coarse levels. We utilize state-of-the-art 3D mesh generation techniques to effectively implement the method. For 3D crystalline defects, such as vacancies, micro-cracks and dislocations, compared with brute-force optimization, complexity with superior rates can be observed numerically, and the strategy has a five-fold acceleration in terms of CPU time for systems with $10^8$ atoms.

preprint2022arXiv

Adaptive Network Combination for Single-Image Reflection Removal: A Domain Generalization Perspective

Recently, multiple synthetic and real-world datasets have been built to facilitate the training of deep single image reflection removal (SIRR) models. Meanwhile, diverse testing sets are also provided with different types of reflection and scenes. However, the non-negligible domain gaps between training and testing sets make it difficult to learn deep models generalizing well to testing images. The diversity of reflections and scenes further makes it a mission impossible to learn a single model being effective to all testing sets and real-world reflections. In this paper, we tackle these issues by learning SIRR models from a domain generalization perspective. Particularly, for each source set, a specific SIRR model is trained to serve as a domain expert of relevant reflection types. For a given reflection-contaminated image, we present a reflection type-aware weighting (RTAW) module to predict expert-wise weights. RTAW can then be incorporated with adaptive network combination (AdaNEC) for handling different reflection types and scenes, i.e., generalizing to unknown domains. Two representative AdaNEC methods, i.e., output fusion (OF) and network interpolation (NI), are provided by considering both adaptation levels and efficiency. For images from one source set, we train RTAW to only predict expert-wise weights of other domain experts for improving generalization ability, while the weights of all experts are predicted and employed during testing. An in-domain expert (IDE) loss is presented for training RTAW. Extensive experiments show the appealing performance gain of our AdaNEC on different state-of-the-art SIRR networks. Source code and pre-trained models will available at https://github.com/csmliu/AdaNEC.

preprint2022arXiv

Adversarial Examples for Good: Adversarial Examples Guided Imbalanced Learning

Adversarial examples are inputs for machine learning models that have been designed by attackers to cause the model to make mistakes. In this paper, we demonstrate that adversarial examples can also be utilized for good to improve the performance of imbalanced learning. We provide a new perspective on how to deal with imbalanced data: adjust the biased decision boundary by training with Guiding Adversarial Examples (GAEs). Our method can effectively increase the accuracy of minority classes while sacrificing little accuracy on majority classes. We empirically show, on several benchmark datasets, our proposed method is comparable to the state-of-the-art method. To our best knowledge, we are the first to deal with imbalanced learning with adversarial examples.

preprint2022arXiv

Amplitude analysis and branching fraction measurement of the decay $D_{s}^{+} \to K^+π^+π^-$

Using $6.32$ fb$^{-1}$ of $e^{+}e^{-}$ collision data collected at the center-of-mass energies between 4.178 and 4.226 GeV with the BESIII detector, we perform an amplitude analysis of the decay $D^+_s \to K^+π^+π^-$ and determine the amplitudes of the various intermediate states. The absolute branching fraction of $D^+_s\to K^+π^+π^-$ is measured to be ($6.11\pm0.18_{\rm stat.}\pm0.11_{\rm syst.})\times 10^{-3}$. The branching fractions of the dominant intermediate processes $D_{s}^{+} \to K^+ρ^0, ρ^0 \to π^+π^-$ and $D_{s}^{+} \to K^*(892)^0π^+, K^*(892)^0 \to K^+π^-$ are determined to be $(1.96\pm0.19_{\rm stat.}\pm0.23_{\rm syst.})\times 10^{-3}$ and $(1.85\pm0.12_{\rm stat.}\pm0.13_{\rm syst.})\times 10^{-3}$, respectively. The intermediate resonances $f_0(500)$, $f_0(980)$, and $f_0(1370)$ are observed for the first time in this channel.

preprint2022arXiv

Amplitude analysis and branching-fraction measurement of $D_{s}^{+} \to π^{+}π^{0}η^{\prime}$

Using data collected with the BESIII detector in $e^+e^-$ collisions at center-of-mass energies between 4.178 and 4.226 GeV and corresponding to 6.32~fb$^{-1}$ of integrated luminosity, we report the amplitude analysis and branching-fraction measurement of the $D^+_s \to π^+ π^0 η^{\prime}$ decay. We find that the dominant intermediate process is $D^+_s \toρ^+ η^{\prime}$ and the significances of other resonant and nonresonant processes are all less than $3σ$. The upper limits on the branching fractions of $S$-wave and $P$-wave nonresonant components are set to $0.10\%$ and $0.74\%$ at the $90\%$ confidence level, respectively. In addition, the branching fraction of the $D^+_s \to π^+ π^0 η^{\prime}$ decay is measured to be $(6.15\pm0.25(\rm stat.)\pm0.18(\rm syst.))\%$, which receives significant contribution only from $D_s^+\to ρ^+η^{\prime}$ according to the amplitude analysis.

preprint2022arXiv

Are Transformers Effective for Time Series Forecasting?

Recently, there has been a surge of Transformer-based solutions for the long-term time series forecasting (LTSF) task. Despite the growing performance over the past few years, we question the validity of this line of research in this work. Specifically, Transformers is arguably the most successful solution to extract the semantic correlations among the elements in a long sequence. However, in time series modeling, we are to extract the temporal relations in an ordered set of continuous points. While employing positional encoding and using tokens to embed sub-series in Transformers facilitate preserving some ordering information, the nature of the \emph{permutation-invariant} self-attention mechanism inevitably results in temporal information loss. To validate our claim, we introduce a set of embarrassingly simple one-layer linear models named LTSF-Linear for comparison. Experimental results on nine real-life datasets show that LTSF-Linear surprisingly outperforms existing sophisticated Transformer-based LTSF models in all cases, and often by a large margin. Moreover, we conduct comprehensive empirical studies to explore the impacts of various design elements of LTSF models on their temporal relation extraction capability. We hope this surprising finding opens up new research directions for the LTSF task. We also advocate revisiting the validity of Transformer-based solutions for other time series analysis tasks (e.g., anomaly detection) in the future. Code is available at: \url{https://github.com/cure-lab/LTSF-Linear}.

preprint2022arXiv

Atomistic fracture in bcc iron revealed by active learning of Gaussian approximation potential

The prediction of atomistic fracture mechanisms in body-centred cubic (bcc) iron is essential for understanding its semi-brittle nature. Existing atomistic simulations of the crack-tip deformation mechanisms under mode-I loading based on classical interatomic potentials yield contradicting predictions. To enable fracture prediction with quantum accuracy, we develop a Gaussian approximation potential (GAP) using an active learning strategy by extending a density functional theory (DFT) database of ferromagnetic bcc iron. We apply the active learning algorithm and obtain a Fe GAP model with a maximum predicted error of 8 meV/atom over a broad range of stress intensity factors (SIFs) and for four crack systems. The learning efficiency of the approach is analysed, and the predicted critical SIFs are compared with Griffith and Rice theories. The simulations reveal that cleavage along the original crack plane is the crack tip mechanism for {100} and {110} crack planes at T=0K, thus settling a long-standing dispute. Our work also highlights the need for a multiscale approach to predicting fracture and intrinsic ductility, whereby finite temperature, finite loading rate effects and pre-existing defects (e.g. nanovoids, dislocations) should be taken explicitly into account.

preprint2022arXiv

Auggie: Encouraging Effortful Communication through Handcrafted Digital Experiences

Digital communication is often brisk and automated. From auto-completed messages to "likes," research has shown that such lightweight interactions can affect perceptions of authenticity and closeness. On the other hand, effort in relationships can forge emotional bonds by conveying a sense of caring and is essential in building and maintaining relationships. To explore effortful communication, we designed and evaluated Auggie, an iOS app that encourages partners to create digitally handcrafted Augmented Reality (AR) experiences for each other. Auggie is centered around crafting a 3D character with photos, animated movements, drawings, and audio for someone else. We conducted a two-week-long field study with 30 participants (15 pairs), who used Auggie with their partners remotely. Our qualitative findings show that Auggie participants engaged in meaningful effort through the handcrafting process, and felt closer to their partners, although the tool may not be appropriate in all situations. We discuss design implications and future directions for systems that encourage effortful communication.

preprint2022arXiv

Auto Machine Learning for Medical Image Analysis by Unifying the Search on Data Augmentation and Neural Architecture

Automated data augmentation, which aims at engineering augmentation policy automatically, recently draw a growing research interest. Many previous auto-augmentation methods utilized a Density Matching strategy by evaluating policies in terms of the test-time augmentation performance. In this paper, we theoretically and empirically demonstrated the inconsistency between the train and validation set of small-scale medical image datasets, referred to as in-domain sampling bias. Next, we demonstrated that the in-domain sampling bias might cause the inefficiency of Density Matching. To address the problem, an improved augmentation search strategy, named Augmented Density Matching, was proposed by randomly sampling policies from a prior distribution for training. Moreover, an efficient automatical machine learning(AutoML) algorithm was proposed by unifying the search on data augmentation and neural architecture. Experimental results indicated that the proposed methods outperformed state-of-the-art approaches on MedMNIST, a pioneering benchmark designed for AutoML in medical image analysis.

preprint2022arXiv

Axiomatic Explanations for Visual Search, Retrieval, and Similarity Learning

Visual search, recommendation, and contrastive similarity learning power technologies that impact billions of users worldwide. Modern model architectures can be complex and difficult to interpret, and there are several competing techniques one can use to explain a search engine's behavior. We show that the theory of fair credit assignment provides a $\textit{unique}$ axiomatic solution that generalizes several existing recommendation- and metric-explainability techniques in the literature. Using this formalism, we show when existing approaches violate "fairness" and derive methods that sidestep these shortcomings and naturally handle counterfactual information. More specifically, we show existing approaches implicitly approximate second-order Shapley-Taylor indices and extend CAM, GradCAM, LIME, SHAP, SBSM, and other methods to search engines. These extensions can extract pairwise correspondences between images from trained $\textit{opaque-box}$ models. We also introduce a fast kernel-based method for estimating Shapley-Taylor indices that require orders of magnitude fewer function evaluations to converge. Finally, we show that these game-theoretic measures yield more consistent explanations for image similarity architectures.

preprint2022arXiv

Beyond a Video Frame Interpolator: A Space Decoupled Learning Approach to Continuous Image Transition

Video frame interpolation (VFI) aims to improve the temporal resolution of a video sequence. Most of the existing deep learning based VFI methods adopt off-the-shelf optical flow algorithms to estimate the bidirectional flows and interpolate the missing frames accordingly. Though having achieved a great success, these methods require much human experience to tune the bidirectional flows and often generate unpleasant results when the estimated flows are not accurate. In this work, we rethink the VFI problem and formulate it as a continuous image transition (CIT) task, whose key issue is to transition an image from one space to another space continuously. More specifically, we learn to implicitly decouple the images into a translatable flow space and a non-translatable feature space. The former depicts the translatable states between the given images, while the later aims to reconstruct the intermediate features that cannot be directly translated. In this way, we can easily perform image interpolation in the flow space and intermediate image synthesis in the feature space, obtaining a CIT model. The proposed space decoupled learning (SDL) approach is simple to implement, while it provides an effective framework to a variety of CIT problems beyond VFI, such as style transfer and image morphing. Our extensive experiments on a variety of CIT tasks demonstrate the superiority of SDL to existing methods. The source code and models can be found at \url{https://github.com/yangxy/SDL}.

preprint2022arXiv

Blind Image Super-resolution with Elaborate Degradation Modeling on Noise and Kernel

While researches on model-based blind single image super-resolution (SISR) have achieved tremendous successes recently, most of them do not consider the image degradation sufficiently. Firstly, they always assume image noise obeys an independent and identically distributed (i.i.d.) Gaussian or Laplacian distribution, which largely underestimates the complexity of real noise. Secondly, previous commonly-used kernel priors (e.g., normalization, sparsity) are not effective enough to guarantee a rational kernel solution, and thus degenerates the performance of subsequent SISR task. To address the above issues, this paper proposes a model-based blind SISR method under the probabilistic framework, which elaborately models image degradation from the perspectives of noise and blur kernel. Specifically, instead of the traditional i.i.d. noise assumption, a patch-based non-i.i.d. noise model is proposed to tackle the complicated real noise, expecting to increase the degrees of freedom of the model for noise representation. As for the blur kernel, we novelly construct a concise yet effective kernel generator, and plug it into the proposed blind SISR method as an explicit kernel prior (EKP). To solve the proposed model, a theoretically grounded Monte Carlo EM algorithm is specifically designed. Comprehensive experiments demonstrate the superiority of our method over current state-of-the-arts on synthetic and real datasets. The source code is available at https://github.com/zsyOAOA/BSRDM.

preprint2022arXiv

Box-supervised Instance Segmentation with Level Set Evolution

In contrast to the fully supervised methods using pixel-wise mask labels, box-supervised instance segmentation takes advantage of the simple box annotations, which has recently attracted a lot of research attentions. In this paper, we propose a novel single-shot box-supervised instance segmentation approach, which integrates the classical level set model with deep neural network delicately. Specifically, our proposed method iteratively learns a series of level sets through a continuous Chan-Vese energy-based function in an end-to-end fashion. A simple mask supervised SOLOv2 model is adapted to predict the instance-aware mask map as the level set for each instance. Both the input image and its deep features are employed as the input data to evolve the level set curves, where a box projection function is employed to obtain the initial boundary. By minimizing the fully differentiable energy function, the level set for each instance is iteratively optimized within its corresponding bounding box annotation. The experimental results on four challenging benchmarks demonstrate the leading performance of our proposed approach to robust instance segmentation in various scenarios. The code is available at: https://github.com/LiWentomng/boxlevelset.

preprint2022arXiv

CausalMTA: Eliminating the User Confounding Bias for Causal Multi-touch Attribution

Multi-touch attribution (MTA), aiming to estimate the contribution of each advertisement touchpoint in conversion journeys, is essential for budget allocation and automatically advertising. Existing methods first train a model to predict the conversion probability of the advertisement journeys with historical data and calculate the attribution of each touchpoint using counterfactual predictions. An assumption of these works is the conversion prediction model is unbiased, i.e., it can give accurate predictions on any randomly assigned journey, including both the factual and counterfactual ones. Nevertheless, this assumption does not always hold as the exposed advertisements are recommended according to user preferences. This confounding bias of users would lead to an out-of-distribution (OOD) problem in the counterfactual prediction and cause concept drift in attribution. In this paper, we define the causal MTA task and propose CausalMTA to eliminate the influence of user preferences. It systemically eliminates the confounding bias from both static and dynamic preferences to learn the conversion prediction model using historical data. We also provide a theoretical analysis to prove CausalMTA can learn an unbiased prediction model with sufficient data. Extensive experiments on both public datasets and the impression data in an e-commerce company show that CausalMTA not only achieves better prediction performance than the state-of-the-art method but also generates meaningful attribution credits across different advertising channels.

preprint2022arXiv

Class-Balanced Pixel-Level Self-Labeling for Domain Adaptive Semantic Segmentation

Domain adaptive semantic segmentation aims to learn a model with the supervision of source domain data, and produce satisfactory dense predictions on unlabeled target domain. One popular solution to this challenging task is self-training, which selects high-scoring predictions on target samples as pseudo labels for training. However, the produced pseudo labels often contain much noise because the model is biased to source domain as well as majority categories. To address the above issues, we propose to directly explore the intrinsic pixel distributions of target domain data, instead of heavily relying on the source domain. Specifically, we simultaneously cluster pixels and rectify pseudo labels with the obtained cluster assignments. This process is done in an online fashion so that pseudo labels could co-evolve with the segmentation model without extra training rounds. To overcome the class imbalance problem on long-tailed categories, we employ a distribution alignment technique to enforce the marginal class distribution of cluster assignments to be close to that of pseudo labels. The proposed method, namely Class-balanced Pixel-level Self-Labeling (CPSL), improves the segmentation performance on target domain over state-of-the-arts by a large margin, especially on long-tailed categories.

preprint2022arXiv

Classification and a priori estimates for the singular prescribing $Q$-curvature equation on 4-manifold

On $(M,g)$ a compact riemannian $4-$manifold we consider the prescribed $Q-$curvature equation defined on $M$ with finite singular sources. We first prove a classification theorem for singular Liouville equations defined on $\mathbb R^4$ and perform a concentration compactness analysis. Then we derive a quantization result for bubbling solutions and establish a priori estimate under the assumption that certain conformal invariant does not take some quantized values. Furthermore we prove a spherical Harnack inequality around singular sources provided their strength is not an integer. Such an inequality implies that in this case singular sources are \emph{isolated simple blow up points}.

preprint2022arXiv

Convergence analysis of discrete high-index saddle dynamics

Saddle dynamics is a time continuous dynamics to efficiently compute the any-index saddle points and construct the solution landscape. In practice, the saddle dynamics needs to be discretized for numerical computations, while the corresponding numerical analysis are rarely studied in the literature, especially for the high-index cases. In this paper we propose the convergence analysis of discrete high-index saddle dynamics. To be specific, we prove the local linear convergence rates of numerical schemes of high-index saddle dynamics, which indicates that the local curvature in the neighborhood of the saddle point and the accuracy of computing the eigenfunctions are main factors that affect the convergence of discrete saddle dynamics. The proved results serve as compensations for the convergence analysis of high-index saddle dynamics and are substantiated by numerical experiments.

preprint2022arXiv

Counterexamples to Fujita's conjecture on surfaces in positive characteristic

We present counterexamples to Fujita's conjecture in positive characteristics. Precisely, we show that over any algebraically closed field $k$ of characteristic $p>0$ and for any positive integer $m$, there exists a smooth projective surface $S$ with an ample Cartier divisor $A$ such that the adjoint linear system $|K_S+mA|$ is not free of base point. Our surface $S$ is a certain kind of generalization of Raynaud surfaces.

preprint2022arXiv

Cross section measurements of the processes $e^+e^- \rightarrow ωπ^{0}$ and $ωη$ at center-of-mass energies between 3.773 and 4.701 GeV

The Born cross sections of the processes $e^+e^- \rightarrow ωπ^{0}$ and $e^+e^- \rightarrow ωη$ are measured at center-of-mass energies between 3.773 and 4.701 GeV using a total integrated luminosity of 22.7 fb$^{-1}$ collected with the BESIII detector operating at the BEPCII collider. A simple $s^{-n}$ dependence for the continuum process can describe the measured Born cross sections. No significant contributions from the $ψ(4160)$, $Y(4230)$, $Y(4360)$, $ψ(4415)$, $Y(4660)$ resonances are found, which indicates relative small branching fractions for these resonances into the $ωπ^{0}$ and $ωη$ final states.

preprint2022arXiv

DAB-DETR: Dynamic Anchor Boxes are Better Queries for DETR

We present in this paper a novel query formulation using dynamic anchor boxes for DETR (DEtection TRansformer) and offer a deeper understanding of the role of queries in DETR. This new formulation directly uses box coordinates as queries in Transformer decoders and dynamically updates them layer-by-layer. Using box coordinates not only helps using explicit positional priors to improve the query-to-feature similarity and eliminate the slow training convergence issue in DETR, but also allows us to modulate the positional attention map using the box width and height information. Such a design makes it clear that queries in DETR can be implemented as performing soft ROI pooling layer-by-layer in a cascade manner. As a result, it leads to the best performance on MS-COCO benchmark among the DETR-like detection models under the same setting, e.g., AP 45.7\% using ResNet50-DC5 as backbone trained in 50 epochs. We also conducted extensive experiments to confirm our analysis and verify the effectiveness of our methods. Code is available at \url{https://github.com/SlongLiu/DAB-DETR}.

preprint2022arXiv

Dense Learning based Semi-Supervised Object Detection

Semi-supervised object detection (SSOD) aims to facilitate the training and deployment of object detectors with the help of a large amount of unlabeled data. Though various self-training based and consistency-regularization based SSOD methods have been proposed, most of them are anchor-based detectors, ignoring the fact that in many real-world applications anchor-free detectors are more demanded. In this paper, we intend to bridge this gap and propose a DenSe Learning (DSL) based anchor-free SSOD algorithm. Specifically, we achieve this goal by introducing several novel techniques, including an Adaptive Filtering strategy for assigning multi-level and accurate dense pixel-wise pseudo-labels, an Aggregated Teacher for producing stable and precise pseudo-labels, and an uncertainty-consistency-regularization term among scales and shuffled patches for improving the generalization capability of the detector. Extensive experiments are conducted on MS-COCO and PASCAL-VOC, and the results show that our proposed DSL method records new state-of-the-art SSOD performance, surpassing existing methods by a large margin. Codes can be found at \textcolor{blue}{https://github.com/chenbinghui1/DSL}.

preprint2022arXiv

Details or Artifacts: A Locally Discriminative Learning Approach to Realistic Image Super-Resolution

Single image super-resolution (SISR) with generative adversarial networks (GAN) has recently attracted increasing attention due to its potentials to generate rich details. However, the training of GAN is unstable, and it often introduces many perceptually unpleasant artifacts along with the generated details. In this paper, we demonstrate that it is possible to train a GAN-based SISR model which can stably generate perceptually realistic details while inhibiting visual artifacts. Based on the observation that the local statistics (e.g., residual variance) of artifact areas are often different from the areas of perceptually friendly details, we develop a framework to discriminate between GAN-generated artifacts and realistic details, and consequently generate an artifact map to regularize and stabilize the model training process. Our proposed locally discriminative learning (LDL) method is simple yet effective, which can be easily plugged in off-the-shelf SISR methods and boost their performance. Experiments demonstrate that LDL outperforms the state-of-the-art GAN based SISR methods, achieving not only higher reconstruction accuracy but also superior perceptual quality on both synthetic and real-world datasets. Codes and models are available at https://github.com/csjliang/LDL.

preprint2022arXiv

DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detection

We present DINO (\textbf{D}ETR with \textbf{I}mproved de\textbf{N}oising anch\textbf{O}r boxes), a state-of-the-art end-to-end object detector. % in this paper. DINO improves over previous DETR-like models in performance and efficiency by using a contrastive way for denoising training, a mixed query selection method for anchor initialization, and a look forward twice scheme for box prediction. DINO achieves $49.4$AP in $12$ epochs and $51.3$AP in $24$ epochs on COCO with a ResNet-50 backbone and multi-scale features, yielding a significant improvement of $\textbf{+6.0}$\textbf{AP} and $\textbf{+2.7}$\textbf{AP}, respectively, compared to DN-DETR, the previous best DETR-like model. DINO scales well in both model size and data size. Without bells and whistles, after pre-training on the Objects365 dataset with a SwinL backbone, DINO obtains the best results on both COCO \texttt{val2017} ($\textbf{63.2}$\textbf{AP}) and \texttt{test-dev} (\textbf{$\textbf{63.3}$AP}). Compared to other models on the leaderboard, DINO significantly reduces its model size and pre-training data size while achieving better results. Our code will be available at \url{https://github.com/IDEACVR/DINO}.

preprint2022arXiv

Discretization and index-robust error analysis for constrained high-index saddle dynamics on high-dimensional sphere

We develop and analyze numerical discretization to the constrained high-index saddle dynamics, the dynamics searching for the high-index saddle points confined on the high-dimensional unit sphere. Compared with the saddle dynamics without constraints, the constrained high-index saddle dynamics has more complex dynamical forms, and additional operations such as the retraction and vector transport are required due to the constraint, which significantly complicate the numerical scheme and the corresponding numerical analysis. Furthermore, as the existing numerical analysis results usually depend on the index of the saddle points implicitly, the proved numerical accuracy may be reduced if the index is high in many applications, which indicates the lack of robustness with respect to the index. To address these issues, we derive the error estimates for numerical discretization of the constrained high-index saddle dynamics on high-dimensional sphere, and then improve it by providing an index-robust error analysis in an averaged norm by adjusting the relaxation parameters. The developed results provide mathematical supports for the accuracy of numerical computations.

preprint2022arXiv

DP-PSI: Private and Secure Set Intersection

One way to classify private set intersection (PSI) for secure 2-party computation is whether the intersection is (a) revealed to both parties or (b) hidden from both parties while only the computing function of the matched payload is exposed. Both aim to provide cryptographic security while avoiding exposing the unmatched elements of the other. They may, however, be insufficient to achieve security and privacy in one practical scenario: when the intersection is required and the information leaked through the function's output must be considered for legal, ethical, and competitive reasons. Two parties, such as the advertiser and the ads supplier, hold sets of users for PSI computation, for example, to reveal common users to the ads supplier in joint marketing applications. In addition to the security guarantees required by standard PSIs to secure unmatched elements, neither party is allowed to "single out" whether an element/user belongs to the other party or not, even though common users are required for joint advertising. This is a fascinating problem for which none of the PSI techniques have provided a solution. In light of this shortcoming, we compose differential privacy (DP) and S2PC to provide the best of both worlds and propose differentially-private PSI (DP-PSI), a new privacy model that shares PSI's strong security protection while adhering to the GDPR's recent formalization of the notion of excluding "signaling out" attacks by each party except with very low probability.

preprint2022arXiv

E2FIF: Push the limit of Binarized Deep Imagery Super-resolution using End-to-end Full-precision Information Flow

Binary neural network (BNN) provides a promising solution to deploy parameter-intensive deep single image super-resolution (SISR) models onto real devices with limited storage and computational resources. To achieve comparable performance with the full-precision counterpart, most existing BNNs for SISR mainly focus on compensating the information loss incurred by binarizing weights and activations in the network through better approximations to the binarized convolution. In this study, we revisit the difference between BNNs and their full-precision counterparts and argue that the key for good generalization performance of BNNs lies on preserving a complete full-precision information flow as well as an accurate gradient flow passing through each binarized convolution layer. Inspired by this, we propose to introduce a full-precision skip connection or its variant over each binarized convolution layer across the entire network, which can increase the forward expressive capability and the accuracy of back-propagated gradient, thus enhancing the generalization performance. More importantly, such a scheme is applicable to any existing BNN backbones for SISR without introducing any additional computation cost. To testify its efficacy, we evaluate it using four different backbones for SISR on four benchmark datasets and report obviously superior performance over existing BNNs and even some 4-bit competitors.

preprint2022arXiv

Efficient and Degradation-Adaptive Network for Real-World Image Super-Resolution

Efficient and effective real-world image super-resolution (Real-ISR) is a challenging task due to the unknown complex degradation of real-world images and the limited computation resources in practical applications. Recent research on Real-ISR has achieved significant progress by modeling the image degradation space; however, these methods largely rely on heavy backbone networks and they are inflexible to handle images of different degradation levels. In this paper, we propose an efficient and effective degradation-adaptive super-resolution (DASR) network, whose parameters are adaptively specified by estimating the degradation of each input image. Specifically, a tiny regression network is employed to predict the degradation parameters of the input image, while several convolutional experts with the same topology are jointly optimized to specify the network parameters via a non-linear mixture of experts. The joint optimization of multiple experts and the degradation-adaptive pipeline significantly extend the model capacity to handle degradations of various levels, while the inference remains efficient since only one adaptively specified network is used for super-resolving the input image. Our extensive experiments demonstrate that the proposed DASR is not only much more effective than existing methods on handling real-world images with different degradation levels but also efficient for easy deployment. Codes, models and datasets are available at https://github.com/csjliang/DASR.

preprint2022arXiv

Efficient Long-Range Attention Network for Image Super-resolution

Recently, transformer-based methods have demonstrated impressive results in various vision tasks, including image super-resolution (SR), by exploiting the self-attention (SA) for feature extraction. However, the computation of SA in most existing transformer based models is very expensive, while some employed operations may be redundant for the SR task. This limits the range of SA computation and consequently the SR performance. In this work, we propose an efficient long-range attention network (ELAN) for image SR. Specifically, we first employ shift convolution (shift-conv) to effectively extract the image local structural information while maintaining the same level of complexity as 1x1 convolution, then propose a group-wise multi-scale self-attention (GMSA) module, which calculates SA on non-overlapped groups of features using different window sizes to exploit the long-range image dependency. A highly efficient long-range attention block (ELAB) is then built by simply cascading two shift-conv with a GMSA module, which is further accelerated by using a shared attention mechanism. Without bells and whistles, our ELAN follows a fairly simple design by sequentially cascading the ELABs. Extensive experiments demonstrate that ELAN obtains even better results against the transformer-based SR models but with significantly less complexity. The source code can be found at https://github.com/xindongzhang/ELAN.

preprint2022arXiv

Error estimates for Euler discretization of high-index saddle dynamics

High-index saddle dynamics provides an effective means to compute the any-index saddle points and construct the solution landscape. In this paper we prove error estimates for Euler discretization of high-index saddle dynamics with respect to the time step size, which remains untreated in the literature. We overcome the main difficulties that lie in the strong nonlinearity of the saddle dynamics and the orthonormalization procedure in the numerical scheme that is uncommon in standard discretization of differential equations. The derived methods are further extended to study the generalized high-index saddle dynamics for non-gradient systems and provide theoretical support for the accuracy of numerical implementations.

preprint2022arXiv

Estimates of bubbling solutions of $SU(3)$ Toda systems at critical parameters-Part 2

In this article we study bubbling solutions of regular $SU(3)$ Toda systems defined on a Riemann surface. There are two major difficulties corresponding to the profile of bubbling solutions: partial blowup phenomenon and bubble accumulation. We prove that when both parameters tend to critical positions, if there is one fully bubbling blowup point, then under one curvature assumption, all the blowup solutions near a blowup point satisfy a spherical Harnack inequality, which completely rules out the bubble-accumulation phenomenon. This fact is crucial for a number of applications.

preprint2022arXiv

Evidence of Spin Frustration in Vanadium Diselenide Monolayer Magnet

Monolayer VSe2, featuring both charge density wave and magnetism phenomena, represents a unique van der Waals magnet in the family of metallic two-dimensional transition-metal dichalcogenides (2D-TMDs). Herein, by means of in-situ microscopic and spectroscopic techniques, including scanning tunneling microscopy/spectroscopy, synchrotron X-ray and angle-resolved photoemission, and X-ray absorption, direct spectroscopic signatures are established, that identify the metallic 1T-phase and vanadium 3d1 electronic configuration in monolayer VSe2 grown on graphite by molecular-beam epitaxy. Element-specific X-ray magnetic circular dichroism, complemented with magnetic susceptibility measurements, further reveals monolayer VSe2 as a frustrated magnet, with its spins exhibiting subtle correlations, albeit in the absence of a long-range magnetic order down to 2 K and up to a 7 T magnetic field. This observation is attributed to the relative stability of the ferromagnetic and antiferromagnetic ground states, arising from its atomic-scale structural features, such as rotational disorders and edges. The results of this study extend the current understanding of metallic 2D-TMDs in the search for exotic low-dimensional quantum phenomena, and stimulate further theoretical and experimental studies on van der Waals monolayer magnets.

preprint2022arXiv

Exact Feature Distribution Matching for Arbitrary Style Transfer and Domain Generalization

Arbitrary style transfer (AST) and domain generalization (DG) are important yet challenging visual learning tasks, which can be cast as a feature distribution matching problem. With the assumption of Gaussian feature distribution, conventional feature distribution matching methods usually match the mean and standard deviation of features. However, the feature distributions of real-world data are usually much more complicated than Gaussian, which cannot be accurately matched by using only the first-order and second-order statistics, while it is computationally prohibitive to use high-order statistics for distribution matching. In this work, we, for the first time to our best knowledge, propose to perform Exact Feature Distribution Matching (EFDM) by exactly matching the empirical Cumulative Distribution Functions (eCDFs) of image features, which could be implemented by applying the Exact Histogram Matching (EHM) in the image feature space. Particularly, a fast EHM algorithm, named Sort-Matching, is employed to perform EFDM in a plug-and-play manner with minimal cost. The effectiveness of our proposed EFDM method is verified on a variety of AST and DG tasks, demonstrating new state-of-the-art results. Codes are available at https://github.com/YBZh/EFDM.

preprint2022arXiv

Few-shot Multi-hop Question Answering over Knowledge Base

KBQA is a task that requires to answer questions by using semantic structured information in knowledge base. Previous work in this area has been restricted due to the lack of large semantic parsing dataset and the exponential growth of searching space with the increasing hops of relation paths. In this paper, we propose an efficient pipeline method equipped with a pre-trained language model. By adopting Beam Search algorithm, the searching space will not be restricted in subgraph of 3 hops. Besides, we propose a data generation strategy, which enables our model to generalize well from few training samples. We evaluate our model on an open-domain complex Chinese Question Answering task CCKS2019 and achieve F1-score of 62.55% on the test dataset. In addition, in order to test the few-shot learning capability of our model, we ramdomly select 10% of the primary data to train our model, the result shows that our model can still achieves F1-score of 58.54%, which verifies the capability of our model to process KBQA task and the advantage in few-shot Learning.

preprint2022arXiv

First Observation of the Semileptonic Decay $Λ_c^+\rightarrow pK^- e^+ν_e$

Using $4.5~\mathrm{fb}^{-1}$ of $e^+e^-$ annihilation data samples collected at the center-of-mass energies ranging from 4.600~GeV to 4.699~GeV with the BESIII detector at the BEPCII collider, a first study of the semileptonic decays $Λ_c^+\rightarrow pK^-e^+ν_e$, $Λ_c^+\rightarrow Λ(1520) e^+ν_e$ and $Λ_c^+\rightarrow Λ(1405) e^+ν_e$ is performed. The $Λ_c^+\rightarrow pK^-e^+ν_e$ decay is observed with a significance of $8.2σ$ and the branching fraction is measured to be $\mathcal{B}(Λ_c^+\rightarrow pK^- e^+ν_e)=(0.88\pm0.17_{\rm stat.}\pm0.07_{\rm syst.})\times 10^{-3}$. We also report evidence of $Λ_c^+\rightarrow Λ(1520)e^+ν_e$ and $Λ_c^+\rightarrow Λ(1405)e^+ν_e$ with significances of $3.3σ$ and $3.2σ$, respectively, and measure $\mathcal B(Λ^+_c\rightarrow Λ(1520)e^+ν_e)=(1.02\pm0.52_{\rm stat.}\pm0.11_{\rm syst.})\times10^{-3}$ and $\mathcal B(Λ^+_c\rightarrow Λ(1405)[\rightarrow pK^-]e^+ν_e)=(0.42\pm0.19_{\rm stat.}\pm0.04_{\rm syst.})\times10^{-3}$. Combining these with the inclusive semileptonic $Λ_c^+$ branching fraction measured by BESIII, the relative fraction is determined to be $[\mathcal{B}(Λ_c^+\rightarrow pK^-e^+ν_e)/\mathcal{B}(Λ_c^+\rightarrow X e^+ν_e)]=(2.1\pm0.4_{\rm stat.}\pm0.2_{\rm syst.})\%$, which provides a clear confirmation that semileptonic $Λ_c^+$ decays are not saturated by the $Λ\ell^+ν_{\ell}$ final state.

preprint2022arXiv

Frequency-dependent polarization of repeating fast radio bursts-implications for their origin

The polarization of fast radio bursts (FRBs), bright astronomical transients, contains crucial information about their environments. We report polarization measurements of five repeating FRBs, the abundant signals of which enable wide-band observations with two telescopes. A clear trend of lower polarization at lower frequencies was found, which can be well characterized by a single parameter rotation-measure-scatter (σRM) and modeled by multi-path scatter. Sources with higher σRM have higher RM magnitude and scattering timescales. The two sources with the most substantial σRM, FRB 20121102A and FRB 20190520B, are associated with a compact persistent radio source. These properties indicate a complex environment near the repeating FRBs, such as a supernova remnant or a pulsar wind nebula, consistent with their arising from young populations.

preprint2022arXiv

Gain and loss induced topological insulating phase in a non Hermitian electrical circuit

There have been considerable efforts devoted to the study of topological phases in certain non-Hermitian systems that possess real eigenfrequencies in the presence of gain and loss. However, it is challenging to experimentally realize such non-Hermitian topological insulators in either quantum or photonic systems, due to the difficulties in introducing controlled gain and loss. On the other hand, the wide choices of active circuit components provide us with unprecedented convenience and flexibility in engineering non-Hermitian topological insulators in electrical circuits. Here, we report experimental realization of a one-dimensional (1D) non-Hermitian topological circuit which exhibits topologically protected edge state purely induced by gain and loss. We show that by tuning the value of the positive/negative resistors in the circuit, our system can switch between different topological phase regions. The topological edge states and interface states are observed at the circuit edge and at the interface between a trivial and nontrivial circuit, which are manifested by a prominent impedance peak at the mid-gap frequency topologically robust to variations of circuit parameters. Our work opens a new gateway towards actively controllable topological systems.

preprint2022arXiv

Grounded Language-Image Pre-training

This paper presents a grounded language-image pre-training (GLIP) model for learning object-level, language-aware, and semantic-rich visual representations. GLIP unifies object detection and phrase grounding for pre-training. The unification brings two benefits: 1) it allows GLIP to learn from both detection and grounding data to improve both tasks and bootstrap a good grounding model; 2) GLIP can leverage massive image-text pairs by generating grounding boxes in a self-training fashion, making the learned representation semantic-rich. In our experiments, we pre-train GLIP on 27M grounding data, including 3M human-annotated and 24M web-crawled image-text pairs. The learned representations demonstrate strong zero-shot and few-shot transferability to various object-level recognition tasks. 1) When directly evaluated on COCO and LVIS (without seeing any images in COCO during pre-training), GLIP achieves 49.8 AP and 26.9 AP, respectively, surpassing many supervised baselines. 2) After fine-tuned on COCO, GLIP achieves 60.8 AP on val and 61.5 AP on test-dev, surpassing prior SoTA. 3) When transferred to 13 downstream object detection tasks, a 1-shot GLIP rivals with a fully-supervised Dynamic Head. Code is released at https://github.com/microsoft/GLIP.

preprint2022arXiv

Image-Adaptive YOLO for Object Detection in Adverse Weather Conditions

Though deep learning-based object detection methods have achieved promising results on the conventional datasets, it is still challenging to locate objects from the low-quality images captured in adverse weather conditions. The existing methods either have difficulties in balancing the tasks of image enhancement and object detection, or often ignore the latent information beneficial for detection. To alleviate this problem, we propose a novel Image-Adaptive YOLO (IA-YOLO) framework, where each image can be adaptively enhanced for better detection performance. Specifically, a differentiable image processing (DIP) module is presented to take into account the adverse weather conditions for YOLO detector, whose parameters are predicted by a small convolutional neural net-work (CNN-PP). We learn CNN-PP and YOLOv3 jointly in an end-to-end fashion, which ensures that CNN-PP can learn an appropriate DIP to enhance the image for detection in a weakly supervised manner. Our proposed IA-YOLO approach can adaptively process images in both normal and adverse weather conditions. The experimental results are very encouraging, demonstrating the effectiveness of our proposed IA-YOLO method in both foggy and low-light scenarios.

preprint2022arXiv

Intelligent Reflecting Surface Networks with Multi-Order-Reflection Effect: System Modelling and Critical Bounds

In this paper, we model, analyze and optimize the multi-user and multi-order-reflection (MUMOR) intelligent reflecting surface (IRS) networks. We first derive a complete MUMOR IRS network model applicable for the arbitrary times of reflections, size and number of IRSs/reflectors. The optimal condition for achieving sum-rate upper bound with one IRS in a closed-form function and the analytical condition to achieve interference-free transmission are derived, respectively. Leveraging this optimal condition, we obtain the MUMOR sum-rate upper bound of the IRS network with different network topologies, where the linear graph (LG), complete graph (CG) and null graph (NG) topologies are considered. Simulation results verify our theories and derivations and demonstrate that the sum-rate upper bounds of different network topologies are under a K-fold improvement given K-piece IRS.

preprint2022arXiv

Introduction to a low-mass dark matter project, ALETHEIA: A Liquid hElium Time projection cHambEr In dArk matter

Dark Matter (DM) is one of the most critical questions to be understood and answered in fundamental physics today. Plenty of astronomical and cosmological observations have already pinned down that DM exists in the Universe, the Milky Way, and the Solar System. However, understanding DM with the language of elementary physics is still in progress. DM direct detection tests the interactive cross-section between galactic DM particles and an underground detector's nucleons. WIMPs is the most discussed DM candidate. After decades of hunting, a convincing WIMPs signal is still at large. Relatively, the low-mass WIMPs region ($\sim$ 10 MeV/c$^2$ - 10 GeV/c$^2$) has not been fully exploited compared to high-mass WIMPs ($\sim$ 10 GeV/c$^2$ - 10 TeV/c$^2$). By filling the arguably cleanest bulk material, LHe, into the arguably most competitive detector in the field, TPCs, ALETHEIA is supposed to achieve an extremely low-level background; therefore, to help answer one of the most pressing physical questions today: the nature of DM. In this paper, we briefly go through the physics motivation of low-mass DM, the ALETHEIA detector's design, possible analysis channels available for DM searches, and the progress we have made since the project launched in the summer of 2020.

preprint2022arXiv

KGRGRL: A User's Permission Reasoning Method Based on Knowledge Graph Reward Guidance Reinforcement Learning

In general, multiple domain cyberspace security assessments can be implemented by reasoning user's permissions. However, while existing methods include some information from the physical and social domains, they do not provide a comprehensive representation of cyberspace. Existing reasoning methods are also based on expert-given rules, resulting in inefficiency and a low degree of intelligence. To address this challenge, we create a Knowledge Graph (KG) of multiple domain cyberspace in order to provide a standard semantic description of the multiple domain cyberspace. Following that, we proposed a user's permissions reasoning method based on reinforcement learning. All permissions in cyberspace are represented as nodes, and an agent is trained to find all permissions that user can have according to user's initial permissions and cyberspace KG. We set 10 reward setting rules based on the features of cyberspace KG in the reinforcement learning of reward information setting, so that the agent can better locate user's all permissions and avoid blindly finding user's permissions. The results of the experiments showed that the proposed method can successfully reason about user's permissions and increase the intelligence level of the user's permissions reasoning method. At the same time, the F1 value of the proposed method is 6% greater than that of the Translating Embedding (TransE) method.

preprint2022arXiv

Large-Scale Pre-training for Person Re-identification with Noisy Labels

This paper aims to address the problem of pre-training for person re-identification (Re-ID) with noisy labels. To setup the pre-training task, we apply a simple online multi-object tracking system on raw videos of an existing unlabeled Re-ID dataset "LUPerson" nd build the Noisy Labeled variant called "LUPerson-NL". Since theses ID labels automatically derived from tracklets inevitably contain noises, we develop a large-scale Pre-training framework utilizing Noisy Labels (PNL), which consists of three learning modules: supervised Re-ID learning, prototype-based contrastive learning, and label-guided contrastive learning. In principle, joint learning of these three modules not only clusters similar examples to one prototype, but also rectifies noisy labels based on the prototype assignment. We demonstrate that learning directly from raw videos is a promising alternative for pre-training, which utilizes spatial and temporal correlations as weak supervision. This simple pre-training task provides a scalable way to learn SOTA Re-ID representations from scratch on "LUPerson-NL" without bells and whistles. For example, by applying on the same supervised Re-ID method MGN, our pre-trained model improves the mAP over the unsupervised pre-training counterpart by 5.7%, 2.2%, 2.3% on CUHK03, DukeMTMC, and MSMT17 respectively. Under the small-scale or few-shot setting, the performance gain is even more significant, suggesting a better transferability of the learned representation. Code is available at https://github.com/DengpanFu/LUPerson-NL

preprint2022arXiv

Learning High-quality Proposals for Acne Detection

Acne detection is crucial for interpretative diagnosis and precise treatment of skin disease. The arbitrary boundary and small size of acne lesions lead to a significant number of poor-quality proposals in two-stage detection. In this paper, we propose a novel head structure for Region Proposal Network to improve the proposals' quality in two ways. At first, a Spatial Aware Double Head(SADH) structure is proposed to disentangle the representation learning for classification and localization from two different spatial perspectives. The proposed SADH ensures a steeper classification confidence gradient and suppresses the proposals having low intersection-over-union(IoU) with the matched ground truth. Then, we propose a Normalized Wasserstein Distance prediction branch to improve the correlation between the proposals' classification scores and IoUs. In addition, to facilitate further research on acne detection, we construct a new dataset named AcneSCU, with high-resolution imageries, precise annotations, and fine-grained lesion categories. Extensive experiments are conducted on both AcneSCU and the public dataset ACNE04, and the results demonstrate the proposed method could improve the proposals' quality, consistently outperforming state-of-the-art approaches. Code and the collected dataset are available in https://github.com/pingguokiller/acnedetection.

preprint2022arXiv

Magnetic Transition in Monolayer VSe2 via Interface Hybridization

Magnetism in monolayer (ML) VSe2 has attracted broad interest in spintronics while existing reports have not reached consensus. Using element-specific X-ray magnetic circular dichroism, a magnetic transition in ML VSe2 has been demonstrated at the contamination-free interface between Co and VSe2. Via interfacial hybridization with Co atomic overlayer, a magnetic moment of about 0.4 uB per V atom in ML VSe2 is revealed, approaching values predicted by previous theoretical calculations. Promotion of the ferromagnetism in ML VSe2 is accompanied by its antiferromagnetic coupling to Co and a reduction in the spin moment of Co. In comparison to the absence of this interface-induced ferromagnetism at the Fe/MLMoSe2 interface, these findings at the Co/ML-VSe2 interface provide clear proof that the ML VSe2, initially with magnetic disorder, is on the verge of magnetic transition.

preprint2022arXiv

Masked Surfel Prediction for Self-Supervised Point Cloud Learning

Masked auto-encoding is a popular and effective self-supervised learning approach to point cloud learning. However, most of the existing methods reconstruct only the masked points and overlook the local geometry information, which is also important to understand the point cloud data. In this work, we make the first attempt, to the best of our knowledge, to consider the local geometry information explicitly into the masked auto-encoding, and propose a novel Masked Surfel Prediction (MaskSurf) method. Specifically, given the input point cloud masked at a high ratio, we learn a transformer-based encoder-decoder network to estimate the underlying masked surfels by simultaneously predicting the surfel positions (i.e., points) and per-surfel orientations (i.e., normals). The predictions of points and normals are supervised by the Chamfer Distance and a newly introduced Position-Indexed Normal Distance in a set-to-set manner. Our MaskSurf is validated on six downstream tasks under three fine-tuning strategies. In particular, MaskSurf outperforms its closest competitor, Point-MAE, by 1.2\% on the real-world dataset of ScanObjectNN under the OBJ-BG setting, justifying the advantages of masked surfel prediction over masked point cloud reconstruction. Codes will be available at https://github.com/YBZh/MaskSurf.

preprint2022arXiv

Mathematical and numerical analysis to shrinking-dimer saddle dynamics with local Lipschitz conditions

We present a mathematical and numerical investigation to the shrinkingdimer saddle dynamics for finding any-index saddle points in the solution landscape. Due to the dimer approximation of Hessian in saddle dynamics, the local Lipschitz assumptions and the strong nonlinearity for the saddle dynamics, it remains challenges for delicate analysis, such as the the boundedness of the solutions and the dimer error. We address these issues to bound the solutions under proper relaxation parameters, based on which we prove the error estimates for numerical discretization to the shrinking-dimer saddle dynamics by matching the dimer length and the time step size. Furthermore, the Richardson extrapolation is employed to obtain a high-order approximation. The inherent reason of requiring the matching of the dimer length and the time step size lies in that the former serves a different mesh size from the later, and thus the proposed numerical method is close to a fully-discrete numerical scheme of some spacetime PDE model with the Hessian in the saddle dynamics and its dimer approximation serving as a "spatial operator" and its discretization, respectively, which in turn indicates the PDE nature of the saddle dynamics.

preprint2022arXiv

Maximizing the Use of Environmental Constraints: A Pushing-Based Hybrid Position/Force Assembly Skill for Contact-Rich Tasks

The need for contact-rich tasks is rapidly growing in modern manufacturing settings. However, few traditional robotic assembly skills consider environmental constraints during task execution, and most of them use these constraints as termination conditions. In this study, we present a pushing-based hybrid position/force assembly skill that can maximize environmental constraints during task execution. To the best of our knowledge, this is the first work that considers using pushing actions during the execution of the assembly tasks. We have proved that our skill can maximize the utilization of environmental constraints using mobile manipulator system assembly task experiments, and achieve a 100\% success rate in the executions.

preprint2022arXiv

Measurement of $e^{+}e^{-} \to K^{+}K^{-}π^{0}$ cross section and observation of a resonant structure

Based on $e^{+}e^{-}$ collision data collected by the BESIII detector at the BEPCII collider at center-of-mass energies from 2.000 to 3.080 GeV, a partial-wave analysis is performed for the process $e^{+}e^{-} \to K^{+}K^{-}π^{0}$. The Born cross section of the process $e^{+}e^{-} \to K^{+}K^{-}π^{0}$ and its subprocesses $e^{+}e^{-} \to ϕπ^{0}$, $K^{*}(892)K$ and $K^{*}_{2}(1430)K$ are measured. The results for $e^{+}e^{-} \to K^{+}K^{-}π^{0}$ and $ϕπ^{0}$ are consistent with the BaBar measurements and with improved precision. By analyzing the cross section, of the subprocesses $e^{+}e^{-} \to$ $K^{*}(892)K$ and $K^{*}_{2}(1430)K$, a structure with mass $M_R$ = (2208 $\pm$ 19 $\pm$ 24) MeV/$c^{2}$ and width $Γ_R$ = (168 $\pm$ 24 $\pm$ 39) MeV is observed with a combined statistical significance of 7.6$σ$. The measured resonance parameters suggest it can be identified as the $ϕ(2170)$, thus the results provide valuable input to understand the internal nature of this state.

preprint2022arXiv

Measurement of $Λ$ baryon polarization in $e^+e^-\rightarrowΛ\barΛ$ at $\sqrt{s} = 3.773$ GeV

Using a data sample of $ψ(3770)$ events collected with the BESIII detector at BEPCII corresponding to an integrated luminosity of 2.9 fb$^{-1}$, we report a measurement of $Λ$ spin polarization in $e^+e^-\rightarrowΛ\barΛ$ at $\sqrt{s} = 3.773$ GeV. The significance of polarization is found to be 2$σ$ including the systematic uncertainty, which implies a zero phase between the transition amplitudes of the $Λ\barΛ$ helicity states. This phase can be interpreted in terms of psionic form factors, and is determined to be $ΔΦ^Ψ$ = $Φ^Ψ_{E} - Φ^Ψ_{M}$ = $(71^{+66}_{-46}$ $\pm$ 5)$^{\circ}$. Similarly, the ratio between the form factors is found to be $R^ψ$ = $|G^Ψ_{E}/G^Ψ_{M}|$ = $0.48^{+0.12}_{-0.07}$ $\pm$ 0.04. The first uncertainties are statistical and the second systematic.

preprint2022arXiv

Measurement of the $D \to K^-π^+π^+π^-$ and $D \to K^-π^+π^0$ coherence factors and average strong-phase differences in quantum-correlated ${D\bar{D}}$ decays

The decays $D\to K^-π^+π^+π^-$ and $D \to K^-π^+π^0$ are studied in a sample of quantum-correlated $D\bar{D}$ pairs produced through the process $e^+e^- \to ψ(3770) \to D\bar{D}$, exploiting a data set collected by the BESIII experiment that corresponds to an integrated luminosity of 2.93 fb$^{-1}$. Here $D$ indicates a quantum superposition of a $D^0$ and a $\bar{D}^0$ meson. By reconstructing one neutral charm meson in a signal decay, and the other in the same or a different final state, observables are measured that contain information on the coherence factors and average strong-phase differences of each of the signal modes. These parameters are critical inputs in the measurement of the angle $γ$ of the Unitarity Triangle in $B^- \to DK^-$ decays at the LHCb and Belle II experiments. The coherence factors are determined to be $R_{K3π}=0.52^{+0.12}_{-0.10}$ and $R_{Kππ^0}=0.78 \pm 0.04$, with values for the average strong-phase differences that are $δ_D^{K3π}=\left(167^{+31}_{-19}\right)^\circ$ and $δ_D^{Kππ^0}=\left(196^{+14}_{-15}\right)^\circ$, where the uncertainties include both statistical and systematic contributions. The analysis is re-performed in four bins of the phase-space of the $D \to K^-π^+π^+π^-$ to yield results that will allow for a more sensitive measurement of $γ$ with this mode, to which the BESIII inputs will contribute an uncertainty of around 6$^\circ$.

preprint2022arXiv

Measurement of the branching fraction and decay asymmetry of $Λ\to nγ$

The radiative hyperon decay $Λ\to nγ$ is studied using $(10087\pm44)\times 10^6$ $J/ψ$ events collected with the BESIII detector operating at BEPCII. The absolute branching fraction of the decay $Λ\to nγ$ is determined with a significance of 5.6$σ$ to be $[0.832\pm0.038(\rm stat.)\pm0.054(\rm syst.)]\times10^{-3}$, which lies significantly below the current PDG value. By analyzing the joint angular distribution of the decay products, the first determination of the decay asymmetry $α_γ$ is reported with a value of $-0.16\pm0.10(\rm stat.)\pm0.05(\rm syst.)$.

preprint2022arXiv

Measurement of the branching fraction for $ψ(3686)\to ωK^0_SK^0_S$

Analyzing $(448.1\pm2.9)\times10^6$ $ψ(3686)$ events collected with the BESIII detector at the BEPCII collider, the $ψ(3686)\to ωK_{S}^{0}K_{S}^{0}$ decay is observed for the first time. The branching fraction for this decay is determined to be $\mathcal{B}_{ψ(3686)\to ωK_{S}^{0}K^{0}_{S}}$=$(7.04\pm0.39\pm0.36)$$\times10^{-5}$, where the first uncertainty is statistical and the second is systematic.

preprint2022arXiv

Measurement of the branching fraction of the doubly Cabibbo-suppressed decay $D^0\to K^+π^-π^0$ and search for $D^0\to K^+π^-π^0π^0$

Using $2.93\,\rm fb^{-1}$ of $e^+e^-$ collision data collected at a center-of-mass energy of 3.773\,GeV with the BESIII detector, we present a measurement of the branching fraction of the doubly Cabibbo-suppressed (DCS) decay $D^0\to K^+π^-π^0$ and a search for the DCS decay $D^0\to K^+π^-π^0π^0$. The branching fraction of $D^0\to K^+π^-π^0$ is determined to be $[3.13^{+0.60}_{-0.56}({\rm stat}) \pm 0.09({\rm syst})] \times 10^{-4}$. No signal is observed for $D^0\to K^+π^-π^0π^0$ and an upper limit of $3.6 \times 10^{-4}$ is set on the branching fraction at the 90\% C.L. We combine these results with the world-average branching fractions of their counterpart Cabibbo-favored decays to determine the ratios of the doubly Cabibbo-suppressed over the Cabibbo-favored branching fractions, ${\mathcal B}(D^0\to K^+π^-π^0)/{\mathcal B}(D^0\to K^-π^+π^0)=(0.22\pm 0.04)\%$~and ${\mathcal B}(D^0\to K^+π^-π^0π^0)/{\mathcal B}(D^0\to K^-π^+π^0π^0)<0.40\%$ at the 90\% C.L., which correspond to $(0.75\pm 0.14)\tan^{4} θ_C$~and $1.37\times \tan^{4} θ_C$, respectively, where $θ_C$ is the Cabibbo angle.

preprint2022arXiv

Measurement of the Cross Section for $e^{+}e^{-}\to$ hadrons at Energies from 2.2324 to 3.6710 GeV

Based on electron-positron collision data collected with the BESIII detector operating at the Beijing Electron Positron Collider II storage rings, the value of $R\equivσ(e^{+}e^{-}\to$hadrons)/$σ(e^{+}e^{-}\toμ^{+}μ^{-})$ is measured at 14 center-of-mass energies from 2.2324 to 3.6710 GeV. The resulting uncertainties are less than $3.0\%$, and are dominated by systematic uncertainties.

preprint2022arXiv

Measurement of the cross section of $e^{+}e^{-}\toηπ^{+}π^{-}$ at center-of-mass energies from 3.872 GeV to 4.700 GeV

Using data samples with an integrated luminosity of 19 fb$^{-1}$ at twenty-eight center-of-mass energies from 3.872 GeV to 4.700 GeV collected with the BESIII detector at the BEPCII electron--positron collider, the process $e^{+}e^{-}\toηπ^{+}π^{-}$ and the intermediate process $e^{+}e^{-}\toηρ^{0}$ are studied for the first time. The Born cross sections are measured. No significant resonance structure is observed in the cross section lineshape.

preprint2022arXiv

Measurement of the total and leptonic decay widths of the $J/ψ$ resonance with an energy scan method at BESIII

Using $e^+e^-$ annihilation data sets collected with the BESIII detector, we measure the cross sections of the processes $e^+e^- \to e^+e^-$ and $e^+e^- \to μ^+μ^-$ at fifteen center-of-mass energy points in the vicinity of the $J/ψ$ resonance. By a simultaneous fit to the measured, center-of-mass energy dependent cross sections of the two processes, the combined quantities $Γ_{ee} Γ_{ee} / Γ_{\rm tot}$ and $Γ_{ee} Γ_{μμ} / Γ_{\rm tot}$ are determined to be ($0.346 \pm 0.009$) and ($0.335 \pm 0.006$) keV, respectively, where $Γ_{ee}$, $Γ_{μμ}$, and $Γ_{\rm tot}$ are the electronic, muonic, and total decay widths of the $J/ψ$ resonance, respectively. Using the resultant $Γ_{ee} Γ_{μμ} / Γ_{\rm tot}$ and $Γ_{ee} Γ_{ee} / Γ_{\rm tot}$, the ratio $Γ_{ee} / Γ_{μμ}$ is calculated to be $1.031 \pm 0.015$, which is consistent with the expectation of lepton universality within about two standard deviations. Assuming lepton universality and using the branching fraction of the $J/ψ$ leptonic decay measured by BESIII in 2013, $Γ_{\rm tot}$ and $Γ_{ll}$ are determined to be ($93.0 \pm 2.1$) and ($5.56 \pm 0.11$) keV, respectively, where $Γ_{ll}$ is the average leptonic decay width of the $J/ψ$ resonance.

preprint2022arXiv

Measurements of Absolute Branching Fractions of $D^0\to K_L^0ϕ$, $K_L^0η$, $K_L^0ω$, and $K_L^0η^{\prime}$

We report the first measurements of the absolute branching fractions of $D^0\to K_L^0ϕ$, $D^0\to K_L^0η$, $D^0\to K_L^0ω$, and $D^0\to K_L^0η^{\prime}$, obtained by analyzing $2.93\,\rm fb^{-1}$ of $e^+e^-$ collision data taken at a center-of-mass energy of 3.773 GeV with the BESIII detector. Taking the world averages of the branching fractions of $D^0\to K_S^0ϕ$, $D^0\to K_S^0η$, $D^0\to K_S^0ω$, and $D^0\to K_S^0η^{\prime}$, the $K_S^0$-$K_L^0$ asymmetry $\mathcal{R}(D^0)$ in these decay modes are obtained. The CP asymmetries in these decays are also determined. No significant $CP$ violation is observed.

preprint2022arXiv

Measurements of the absolute branching fractions of hadronic $D$-meson decays involving kaons and pions

By analyzing an electron-positron collision data sample corresponding to an integrated luminosity of $2.93\,\rm fb^{-1}$ taken at the center-of-mass energy of 3.773 GeV with the BESIII detector, we obtain for the first time the absolute branching fractions for seven $D^0$ and $D^+$ hadronic decay modes and search for the hadronic decay $D^0\to K^0_S K^0_Sπ^0$ with much improved sensitivity. The results are ${\mathcal B}(D^0\to K^0_Sπ^0π^0π^0 )=( 7.64\pm 0.30\pm 0.29)\times 10^{-3}$, ${\mathcal B}(D^0\to K^-π^+π^0π^0π^0 )=( 9.54\pm 0.30\pm 0.31)\times 10^{-3}$, ${\mathcal B}(D^0\to K^0_Sπ^+π^-π^0π^0)=(12.66\pm 0.45\pm 0.43)\times 10^{-3}$, ${\mathcal B}(D^+\to K^0_Sπ^+π^0π^0 )=(29.04\pm 0.62\pm 0.87)\times 10^{-3}$, ${\mathcal B}(D^+\to K^0_Sπ^+π^+π^-π^0)=(15.28\pm 0.57\pm 0.60)\times 10^{-3}$, ${\mathcal B}(D^+\to K^0_Sπ^+π^0π^0π^0)=( 5.54\pm 0.44\pm 0.32)\times 10^{-3}$, ${\mathcal B}(D^+\to K^-π^+π^+π^0π^0 )=( 4.95\pm 0.26\pm 0.19)\times 10^{-3}$, ${\mathcal B}({D^0\to K^0_S K^0_Sπ^0}) < 1.57 \times 10^{-4}$ at the 90\% confidence level. Here the first uncertainties are statistical and the second ones systematic. The newly studied decays greatly enrich the knowledge of the $D\to \bar Kπππ$ and $D\to \bar Kππππ$ hadronic decays, and open a bridge to access more two-body hadronic $D$ decays containing scalar, vector, axial and tensor mesons in the charm sector.

preprint2022arXiv

Metaverse Native Communication: A Blockchain and Spectrum Prospective

Metaverse depicts a vista of constructing a virtual environment parallel to the real world so people can communicate with others and objects through digital entities. In the real world, communication relies on identities and addresses that are recognized by authorities, no matter the link is established via post, email, mobile phone, or landline. Metaverse, however, is different from the real world, which requires a single identity belongs to the individual. This identity can be an encrypted virtual address in the metaverse but no one can trace or verify it. In order to achieve such addresses to hide individuals in the metaverse, re-mapping the virtual address to the individual's identity and a specific spectrum to support the address-based communication for the metaverse are needed. Therefore, metaverse native or meta-native communications based on blockchain could be a promising solution to directly connect entities with their native encrypted addresses that gets rid of the existing network services based on IP, cellular, HTTP, etc. This paper proposes a vision of blockchain, encrypted address and address-based access model for all users, devices, services, etc. to contribute to the metaverse. Furthermore, the allocation architecture of a designated spectrum for the metaverse is proposed to remove the barrier to access to the metaverse/blockchain in response to the initiatives of metaverse and decentralized Internet.

preprint2022arXiv

MorphoSim: An efficient and scalable phase-field framework for accurately simulating multicellular morphologies

The phase field model can accurately simulate the evolution of microstructures with complex morphologies, and it has been widely used for cell modeling in the last two decades. However, compared to other cellular models such as the coarse-grained model and the vertex model, its high computational cost caused by three-dimensional spatial discretization hampered its application and scalability, especially for multicellular organisms. Recently, we built a phase field model coupled with in vivo imaging data to accurately reconstruct the embryonic morphogenesis of Caenorhabditis elegans from 1- to 8-cell stages [Kuang et al, PLoS Comput. Biol., 2022]. In this work, we propose an improved phase field model by using the stabilized numerical scheme and modified volume constriction. Then we present a scalable phase-field framework, MorphoSim, which is 100 times more efficient than the previous one, and can simulate over 100 mechanically interacting cells. Finally, we demonstrate how MorphoSim can be successfully applied to reproduce the assembly, self-repairing, and dissociation of a synthetic artificial multicellular system - the synNotch system.

preprint2022arXiv

Multiple Domain Cyberspace Attack and Defense Game Based on Reward Randomization Reinforcement Learning

The existing network attack and defense method can be regarded as game, but most of the game only involves network domain, not multiple domain cyberspace. To address this challenge, this paper proposed a multiple domain cyberspace attack and defense game model based on reinforcement learning. We define the multiple domain cyberspace include physical domain, network domain and digital domain. By establishing two agents, representing the attacker and the defender respectively, defender will select the multiple domain actions in the multiple domain cyberspace to obtain defender's optimal reward by reinforcement learning. In order to improve the defense ability of defender, a game model based on reward randomization reinforcement learning is proposed. When the defender takes the multiple domain defense action, the reward is randomly given and subject to linear distribution, so as to find the better defense policy and improve defense success rate. The experimental results show that the game model can effectively simulate the attack and defense state of multiple domain cyberspace, and the proposed method has a higher defense success rate than DDPG and DQN.

preprint2022arXiv

Mutual Consistency Learning for Semi-supervised Medical Image Segmentation

In this paper, we propose a novel mutual consistency network (MC-Net+) to effectively exploit the unlabeled data for semi-supervised medical image segmentation. The MC-Net+ model is motivated by the observation that deep models trained with limited annotations are prone to output highly uncertain and easily mis-classified predictions in the ambiguous regions (e.g., adhesive edges or thin branches) for medical image segmentation. Leveraging these challenging samples can make the semi-supervised segmentation model training more effective. Therefore, our proposed MC-Net+ model consists of two new designs. First, the model contains one shared encoder and multiple slightly different decoders (i.e., using different up-sampling strategies). The statistical discrepancy of multiple decoders' outputs is computed to denote the model's uncertainty, which indicates the unlabeled hard regions. Second, we apply a novel mutual consistency constraint between one decoder's probability output and other decoders' soft pseudo labels. In this way, we minimize the discrepancy of multiple outputs (i.e., the model uncertainty) during training and force the model to generate invariant results in such challenging regions, aiming at regularizing the model training. We compared the segmentation results of our MC-Net+ model with five state-of-the-art semi-supervised approaches on three public medical datasets. Extension experiments with two standard semi-supervised settings demonstrate the superior performance of our model over other methods, which sets a new state of the art for semi-supervised medical image segmentation. Our code is released publicly at https://github.com/ycwu1997/MC-Net.

preprint2022arXiv

NTIRE 2021 Multi-modal Aerial View Object Classification Challenge

In this paper, we introduce the first Challenge on Multi-modal Aerial View Object Classification (MAVOC) in conjunction with the NTIRE 2021 workshop at CVPR. This challenge is composed of two different tracks using EO andSAR imagery. Both EO and SAR sensors possess different advantages and drawbacks. The purpose of this competition is to analyze how to use both sets of sensory information in complementary ways. We discuss the top methods submitted for this competition and evaluate their results on our blind test set. Our challenge results show significant improvement of more than 15% accuracy from our current baselines for each track of the competition

preprint2022arXiv

Observation of $a_0(1710)^+ \to K_S^0K^+$ in study of the $D_s^+\to K_S^0K^+π^0$ decay

Using $e^+e^-$ annihilation data corresponding to an integrated luminosity of 6.32 fb$^{-1}$ collected at center-of-mass energies between 4.178 GeV and 4.226 GeV with the BESIII detector, we perform the first amplitude analysis of the decay $D_s^+\to K_S^0K^+π^0$ and determine the relative branching fractions and phases for intermediate processes. We observe the $a_0(1710)^+$, the isovector partner of the $f_0(1710)$ and $f_0(1770)$ mesons, in its decay to $K_S^0K^+$ for the first time. In addition, we measure the ratio $\frac{\mathcal{B}(D_{s}^{+} \to \bar{K}^{*}(892)^{0}K^{+})}{\mathcal{B}(D_{s}^{+} \to \bar{K}^{0}K^{*}(892)^{+})}$ to be $2.35^{+0.42}_{-0.23\text{stat.}}\pm 0.10_{\rm syst.}$. Finally, we provide a precision measurement of the absolute branching fraction $\mathcal{B}(D_s^+\to K_S^0K^+π^0) = (1.46\pm 0.06_{\text{stat.}}\pm 0.05_{\text{syst.}})\%$.

preprint2022arXiv

Observation of $η_c(2S) \to 3(π^+π^-)$ and measurements of $χ_{cJ} \to 3(π^+π^-)$ in $ψ(3686)$ radiative transitions

The hadronic decay $η_c(2S) \to 3(π^+π^-)$ is observed with a statistical significance of 9.3 standard deviations using $(448.1\pm2.9)\times10^6$ $ψ(3686)$ events collected by the BESIII detector at the BEPCII collider. The measured mass and width of $η_c(2S)$ are $(3643.4 \pm 2.3 (\rm stat.) \pm 4.4 (\rm syst.))$ MeV/$c^2$ and $(19.8 \pm 3.9 (\rm stat.) \pm 3.1 (\rm syst.))$ MeV, respectively, which are consistent with the world average values within two standard deviations. The product branching fraction $\mathcal{B}[ψ(3686)\to γη_c(2S)]\times\mathcal{B}[η_c(2S)\to3(π^+π^-)]$ is measured to be $(9.2 \pm 1.0 (\rm stat.) \pm 0.9 (\rm syst.))\times10^{-6}$. Using $\mathcal{B}[ψ(3686)\to γη_c(2S)]=(7.0^{+3.4}_{-2.5})\times10^{-4}$, we obtain $\mathcal{B}[η_c(2S) \to 3(π^+π^-)] = (1.31 \pm 0.15 (\rm stat.) \pm 0.13 (\rm syst.)(^{+0.64}_{-0.47}) (\rm extr))\times10^{-2}$, where the third uncertainty is from $\mathcal{B}[ψ(3686) \to γη_c(2S)]$. We also measure the $χ_{cJ} \to 3(π^+π^-)$ ($J=0, 1, 2$) decays via $ψ(3686) \to γχ_{cJ}$ transitions. The branching fractions are $\mathcal{B}[χ_{c0} \to 3(π^+π^-)] = (2.080\pm0.006 (\rm stat.)\pm0.068 (\rm syst.))\times10^{-2}$, $\mathcal{B}[χ_{c1} \to 3(π^+π^-)] = (1.092\pm0.004 (\rm stat.)\pm0.035 (\rm syst.))\times10^{-2}$, and $\mathcal{B}[χ_{c2} \to 3(π^+π^-)] = (1.565\pm0.005 (\rm stat.)\pm0.048 (\rm syst.))\times10^{-2}$.

preprint2022arXiv

Observation of resonance structures in $e^+e^-\to π^+π^-ψ_2(3823)$ and mass measurement of $ψ_2(3823)$

Using a data sample corresponding to an integrated luminosity of 11.3 $\rm fb^{-1}$ collected at center-of-mass energies from $4.23$ to $4.70$ GeV with the BESIII detector, we measure the product of the $e^+e^-\to π^+π^-ψ_2(3823)$ cross section and the branching fraction $\mathcal{B}[ψ_2(3823)\to γχ_{c1}]$. For the first time, resonance structure is observed in the cross section line shape of $e^+e^-\to π^+π^-ψ_2(3823)$ with significances exceeding $5σ$. A fit to data with two coherent Breit-Wigner resonances modeling the $\sqrt{s}$-dependent cross section yields $M(R_1)=4406.9\pm 17.2\pm 4.5$ MeV/$c^2$, $Γ(R_1)=128.1\pm 37.2\pm 2.3$ MeV, and $M(R_2)=4647.9\pm 8.6\pm 0.8$ MeV/$c^2$, $Γ(R_2)=33.1\pm 18.6\pm 4.1$ MeV. Though weakly disfavored by the data, a single resonance with $M(R)=4417.5\pm26.2\pm3.5$ MeV/$c^2$, $Γ(R)=245\pm48\pm13$ MeV is also possible to interpret data. This observation deepens our understanding of the nature of the vector charmoniumlike states. The mass of the $ψ_2(3823)$ state is measured as $(3823.12\pm 0.43\pm 0.13)$ MeV/$c^2$, which is the most precise measurement to date.

preprint2022arXiv

Observation of the double Dalitz decay $η'\to e^+e^-e^+e^-$

Based on $(10087 \pm 44)\times10^6$ $J/ψ$ events collected with the BESIII detector at BEPCII, the double Dalitz decay $η'\to e^+e^-e^+e^-$ is observed for the first time via the $J/ψ\toγη'$ decay process. The significance is found to be 5.7$σ$ with systematic uncertainties taken into consideration. Its branching fraction is determined to be $\mathcal{B}(η'\to e^+ e^- e^+ e^-) =(4.5\pm1.0(\mathrm{stat.})\pm0.5(\mathrm{sys.})) \times 10^{-6}$.

preprint2022arXiv

Observation of the electromagnetic Dalitz decay $D^{\ast 0}\to D^{0}e^{+}e^{-}$

Based on 3.19 fb$^{-1}$ of $e^+e^-$ collision data accumulated at the center-of-mass energy 4.178 GeV with the BESIII detector operating at the BEPCII collider, the electromagnetic Dalitz decay $D^{\ast 0}\to D^{0}e^{+}e^{-}$ is observed for the first time with a statistical significance of $13.2σ$. The ratio of the branching fraction of $D^{\ast 0}\to D^{0}e^{+}e^{-}$ to that of $D^{\ast 0}\to D^{0} γ$ is measured to be $(11.08\pm0.76\pm0.49)\times 10^{-3}$. By using the world average value of the branching fraction of $D^{\ast 0}\to D^{0} γ$, the branching fraction of $D^{\ast 0}\to D^{0}e^{+}e^{-}$ is determined to be $(3.91\pm0.27\pm0.17\pm0.10)\times 10^{-3}$, where the first uncertainty is statistical, the second systematic and the third external branching fractions.

preprint2022arXiv

Observation of the Singly Cabibbo-Suppressed Decay $Λ_{c}^{+} \to nπ^{+}$

The singly Cabibbo-suppressed decay $Λ_{c}^{+} \to nπ^{+}$ is observed for the first time with a statistical significance of $7.3σ$ by using 3.9 $\mathrm{fb}^{-1}$ of $e^{+}e^{-}$ collision data collected at center-of-mass energies between 4.612 and 4.699 GeV with the BESIII detector at BEPCII. The branching fraction of $Λ_{c}^{+} \to nπ^{+}$ is measured to be $(6.6\pm1.2_{\rm stat}\pm0.4_{\rm syst})\times 10^{-4}$. By taking the upper limit of branching fractions of $Λ_{c}^{+} \to pπ^0$ from the Belle experiment, the ratio of branching fractions between $Λ_{c}^{+} \to nπ^{+}$ and $Λ_{c}^{+} \to pπ^0$ is calculated to be larger than 7.2 at the 90% confidence level, which disagrees with the current predictions of available phenomenological models. In addition, the branching fractions of the Cabibbo-favored decays $Λ_{c}^{+} \to Λπ^{+}$ and $Λ_{c}^{+} \to Σ^{0}π^{+}$ are measured to be $(1.31\pm0.08_{\rm stat}\pm0.05_{\rm syst})\times 10^{-2}$ and $(1.22\pm0.08_{\rm stat}\pm0.07_{\rm syst})\times 10^{-2}$, respectively, which are consistent with previous results.

preprint2022arXiv

On multilevel Monte Carlo methods for deterministic and uncertain hyperbolic systems

In this paper, we evaluate the performance of the multilevel Monte Carlo method (MLMC) for deterministic and uncertain hyperbolic systems, where randomness is introduced either in the modeling parameters or in the approximation algorithms. MLMC is a well known variance reduction method widely used to accelerate Monte Carlo (MC) sampling. However, we demonstrate in this paper that for hyperbolic systems, whether MLMC can achieve a real boost turns out to be delicate. The computational costs of MLMC and MC depend on the interplay among the accuracy (bias) and the computational cost of the numerical method for a single sample, as well as the variances of the sampled MLMC corrections or MC solutions. We characterize three regimes for the MLMC and MC performances using those parameters, and show that MLMC may not accelerate MC and can even have a higher cost when the variances of MC solutions and MLMC corrections are of the same order. Our studies are carried out by a few prototype hyperbolic systems: a linear scalar equation, the Euler and shallow water equations, and a linear relaxation model, the above statements are proved analytically in some cases, and demonstrated numerically for the cases of the stochastic hyperbolic equations driven by white noise parameters and Glimm's random choice method for deterministic hyperbolic equations.

preprint2022arXiv

One-stage Video Instance Segmentation: From Frame-in Frame-out to Clip-in Clip-out

Many video instance segmentation (VIS) methods partition a video sequence into individual frames to detect and segment objects frame by frame. However, such a frame-in frame-out (FiFo) pipeline is ineffective to exploit the temporal information. Based on the fact that adjacent frames in a short clip are highly coherent in content, we propose to extend the one-stage FiFo framework to a clip-in clip-out (CiCo) one, which performs VIS clip by clip. Specifically, we stack FPN features of all frames in a short video clip to build a spatio-temporal feature cube, and replace the 2D conv layers in the prediction heads and the mask branch with 3D conv layers, forming clip-level prediction heads (CPH) and clip-level mask heads (CMH). Then the clip-level masks of an instance can be generated by feeding its box-level predictions from CPH and clip-level features from CMH into a small fully convolutional network. A clip-level segmentation loss is proposed to ensure that the generated instance masks are temporally coherent in the clip. The proposed CiCo strategy is free of inter-frame alignment, and can be easily embedded into existing FiFo based VIS approaches. To validate the generality and effectiveness of our CiCo strategy, we apply it to two representative FiFo methods, Yolact \cite{bolya2019yolact} and CondInst \cite{tian2020conditional}, resulting in two new one-stage VIS models, namely CiCo-Yolact and CiCo-CondInst, which achieve 37.1/37.3\%, 35.2/35.4\% and 17.2/18.0\% mask AP using the ResNet50 backbone, and 41.8/41.4\%, 38.0/38.9\% and 18.0/18.2\% mask AP using the Swin Transformer tiny backbone on YouTube-VIS 2019, 2021 and OVIS valid sets, respectively, recording new state-of-the-arts. Code and video demos of CiCo can be found at \url{https://github.com/MinghanLi/CiCo}.

preprint2022arXiv

Online Multi-Object Tracking with Unsupervised Re-Identification Learning and Occlusion Estimation

Occlusion between different objects is a typical challenge in Multi-Object Tracking (MOT), which often leads to inferior tracking results due to the missing detected objects. The common practice in multi-object tracking is re-identifying the missed objects after their reappearance. Though tracking performance can be boosted by the re-identification, the annotation of identity is required to train the model. In addition, such practice of re-identification still can not track those highly occluded objects when they are missed by the detector. In this paper, we focus on online multi-object tracking and design two novel modules, the unsupervised re-identification learning module and the occlusion estimation module, to handle these problems. Specifically, the proposed unsupervised re-identification learning module does not require any (pseudo) identity information nor suffer from the scalability issue. The proposed occlusion estimation module tries to predict the locations where occlusions happen, which are used to estimate the positions of missed objects by the detector. Our study shows that, when applied to state-of-the-art MOT methods, the proposed unsupervised re-identification learning is comparable to supervised re-identification learning, and the tracking performance is further improved by the proposed occlusion estimation module.

preprint2022arXiv

OTExtSum: Extractive Text Summarisation with Optimal Transport

Extractive text summarisation aims to select salient sentences from a document to form a short yet informative summary. While learning-based methods have achieved promising results, they have several limitations, such as dependence on expensive training and lack of interpretability. Therefore, in this paper, we propose a novel non-learning-based method by for the first time formulating text summarisation as an Optimal Transport (OT) problem, namely Optimal Transport Extractive Summariser (OTExtSum). Optimal sentence extraction is conceptualised as obtaining an optimal summary that minimises the transportation cost to a given document regarding their semantic distributions. Such a cost is defined by the Wasserstein distance and used to measure the summary's semantic coverage of the original document. Comprehensive experiments on four challenging and widely used datasets - MultiNews, PubMed, BillSum, and CNN/DM demonstrate that our proposed method outperforms the state-of-the-art non-learning-based methods and several recent learning-based methods in terms of the ROUGE metric.

preprint2022arXiv

Partial wave analysis of $J/ψ\to γη^{\prime} η^{\prime}$

Using a sample of $(10.09~\pm~0.04)\times10^{9} ~J/ψ$ events collected with the BESIII detector, a partial wave analysis of $J/ψ\toγη^{\prime}η^{\prime}$ is performed. The masses and widths of the observed resonances and their branching fractions are reported. The main contribution is from $J/ψ\rightarrowγf_0(2020)$ with $f_0(2020)\rightarrowη^{\prime}η^{\prime}$, which is found with a significance of greater than 25$σ$. The product branching fraction ${\cal B}\left(J/ψ\rightarrowγf_0(2020)\right)\cdot{\cal B}\left(f_0(2020)\rightarrowη^{\prime}η^{\prime}\right)$ is measured to be $(2.63\pm0.06({\rm stat.})^{+0.31}_{-0.46}({\rm syst.}))\times10^{-4}$.

preprint2022arXiv

Radio detection of an elusive millisecond pulsar in the Globular Cluster NGC 6397

We report the discovery of a new 5.78 ms-period millisecond pulsar (MSP), PSR J1740-5340B (NGC 6397B), in an eclipsing binary system discovered with the Parkes radio telescope (now also known as Murriyang), Australia, and confirmed with the MeerKAT radio telescope in South Africa. The measured orbital period, 1.97 days, is the longest among all eclipsing binaries in globular clusters (GCs) and consistent with that of the coincident X-ray source U18, previously suggested to be a 'hidden MSP'. Our XMM-Newton observations during NGC 6397B's radio quiescent epochs detected no X-ray flares. NGC 6397B is either a transitional MSP or an eclipsing binary in its initial stage of mass transfer after the companion star left the main sequence. The discovery of NGC 6397B potentially reveals a subgroup of extremely faint and heavily obscured binary pulsars, thus providing a plausible explanation to the apparent dearth of binary neutron stars in core-collapsed GCs as well as a critical constraint on the evolution of GCs.

preprint2022arXiv

Rapid model transfer for medical image segmentation via iterative human-in-the-loop update: from labelled public to unlabelled clinical datasets for multi-organ segmentation in CT

Despite the remarkable success on medical image analysis with deep learning, it is still under exploration regarding how to rapidly transfer AI models from one dataset to another for clinical applications. This paper presents a novel and generic human-in-the-loop scheme for efficiently transferring a segmentation model from a small-scale labelled dataset to a larger-scale unlabelled dataset for multi-organ segmentation in CT. To achieve this, we propose to use an igniter network which can learn from a small-scale labelled dataset and generate coarse annotations to start the process of human-machine interaction. Then, we use a sustainer network for our larger-scale dataset, and iteratively updated it on the new annotated data. Moreover, we propose a flexible labelling strategy for the annotator to reduce the initial annotation workload. The model performance and the time cost of annotation in each subject evaluated on our private dataset are reported and analysed. The results show that our scheme can not only improve the performance by 19.7% on Dice, but also expedite the cost time of manual labelling from 13.87 min to 1.51 min per CT volume during the model transfer, demonstrating the clinical usefulness with promising potentials.

preprint2022arXiv

Reconfigurable Intelligent Surface-induced Randomness for mmWave Key Generation

Secret key generation in physical layer security exploits the unpredictable random nature of wireless channels. The millimeter-wave (mmWave) channels have limited multipath and channel randomness in static environments. In this paper, for mmWave secret key generation of physical layer security, we use a reconfigurable intelligent surface (RIS) to induce randomness directly in wireless environments, without adding complexity to transceivers. We consider RIS to have continuous individual phase shifts (CIPS) and derive the RIS-assisted reflection channel distribution with its parameters. Then, we propose continuous group phase shifts (CGPS) to increase the randomness specifically at legal parties. Since the continuous phase shifts are expensive to implement, we analyze discrete individual phase shifts (DIPS) and derive the corresponding channel distribution, which is dependent on the quantization bit. We then derive the secret key rate (SKR) to evaluate the randomness performance. With the simulation results verifying the analytical results, this work explains the mathematical principles and lays a foundation for future mmWave evaluation and optimization of artificial channel randomness.

preprint2022arXiv

Recurrent LSTM-based UAV Trajectory Prediction with ADS-B Information

Recently, unmanned aerial vehicles (UAVs) are gathering increasing attentions from both the academia and industry. The ever-growing number of UAV brings challenges for air traffic control (ATC), and thus trajectory prediction plays a vital role in ATC, especially for avoiding collisions among UAVs. However, the dynamic flight of UAV aggravates the complexity of trajectory prediction. Different with civil aviation aircrafts, the most intractable difficulty for UAV trajectory prediction depends on acquiring effective location information. Fortunately, the automatic dependent surveillance-broadcast (ADS-B) is an effective technique to help obtain positioning information. It is widely used in the civil aviation aircraft, due to its high data update frequency and low cost of corresponding ground stations construction. Hence, in this work, we consider leveraging ADS-B to help UAV trajectory prediction. However, with the ADS-B information for a UAV, it still lacks efficient mechanism to predict the UAV trajectory. It is noted that the recurrent neural network (RNN) is available for the UAV trajectory prediction, in which the long short-term memory (LSTM) is specialized in dealing with the time-series data. As above, in this work, we design a system of UAV trajectory prediction with the ADS-B information, and propose the recurrent LSTM (RLSTM) based algorithm to achieve the accurate prediction. Finally, extensive simulations are conducted by Python to evaluate the proposed algorithms, and the results show that the average trajectory prediction error is satisfied, which is in line with expectations.

preprint2022arXiv

Saliency Guided Inter- and Intra-Class Relation Constraints for Weakly Supervised Semantic Segmentation

Weakly supervised semantic segmentation with only image-level labels aims to reduce annotation costs for the segmentation task. Existing approaches generally leverage class activation maps (CAMs) to locate the object regions for pseudo label generation. However, CAMs can only discover the most discriminative parts of objects, thus leading to inferior pixel-level pseudo labels. To address this issue, we propose a saliency guided Inter- and Intra-Class Relation Constrained (I$^2$CRC) framework to assist the expansion of the activated object regions in CAMs. Specifically, we propose a saliency guided class-agnostic distance module to pull the intra-category features closer by aligning features to their class prototypes. Further, we propose a class-specific distance module to push the inter-class features apart and encourage the object region to have a higher activation than the background. Besides strengthening the capability of the classification network to activate more integral object regions in CAMs, we also introduce an object guided label refinement module to take a full use of both the segmentation prediction and the initial labels for obtaining superior pseudo-labels. Extensive experiments on PASCAL VOC 2012 and COCO datasets demonstrate well the effectiveness of I$^2$CRC over other state-of-the-art counterparts. The source codes, models, and data have been made available at \url{https://github.com/NUST-Machine-Intelligence-Laboratory/I2CRC}.

preprint2022arXiv

Search for $X(3872)\toπ^0χ_{c0}$ and $X(3872)\toππχ_{c0}$ at BESIII

Using 9.9 fb$^{-1}$ of $e^+e^-$ collision data collected by the BESIII detector at center-of-mass energies between 4.15 and 4.30 GeV, we search for the processes $e^+e^-\toγX(3872)$ with $X(3872)\rightarrowπ^0χ_{c0}$ and $X(3872)\rightarrowππχ_{c0}$. Depending on the fitting model, the statistical significance for $X(3872)\toπ^0χ_{c0}$ ranges from 1.3$σ$ to 2.8$σ$. We set upper limits (at 90\% C.L.) of $\frac{\mathcal{B}(X(3872)\rightarrowπ^0χ_{c0})}{\mathcal{B}(X(3872)\toπ^+π^-J/ψ)}<3.6$, $\frac{\mathcal{B}(X(3872)\rightarrowπ^+π^-χ_{c0})}{\mathcal{B}(X(3872)\toπ^+π^-J/ψ)}<0.68$, and $\frac{\mathcal{B}(X(3872)\rightarrowπ^0π^0χ_{c0})}{\mathcal{B}(X(3872)\toπ^+π^-J/ψ)}<1.7$. Combined with the BESIII measurement of $X(3872)\toπ^0χ_{c1}$, we also set an upper limit of $\frac{\mathcal{B}(X(3872)\rightarrowπ^0χ_{c0})}{\mathcal{B}(X(3872)\toπ^0χ_{c1})}<4.4$.

preprint2022arXiv

Search for baryon and lepton number violating decays $D^{0}\to \bar{p}e^{+}$ and $D^{0}\to pe^{-}$

Using an electron-positron collision data sample corresponding to an integrated luminosity of 2.93~fb$^{-1}$ collected with the BESIII detector at a center-of-mass energy of 3.773 GeV, we search for the baryon and lepton number violating decays $D^{0}\to \bar{p}e^{+}$ and $D^{0}\to pe^{-}$. No obvious signals are found with the current statistics. The upper limits on the branching fractions for $D^{0}\to \bar{p}e^{+}$ and $D^{0}\to pe^{-}$ are set to be $1.2\times 10^{-6}$ and $2.2\times 10^{-6}$ at 90\% confidence level, respectively.

preprint2022arXiv

Search for baryon and lepton number violation decay $D^{\pm}\to n(\bar{n})e^{\pm}$

Using a data set of electron-positron collisions corresponding to an integrated luminosity of ${\rm 2.93~fb^{-1}}$ taken with the BESIII detector at a center-of-mass energy of 3.773 GeV, a search for the baryon ($B$) and lepton ($L$) number violating decays $D^{\pm}\to n(\bar{n})e^{\pm}$ is performed. No signal is observed and the upper limits on the branching fractions at the $90\%$ confidence level are set to be $1.43\times10^{-5}$ for the decays $D^{+(-)}\to \bar{n}(n)e^{+(-)}$ with $Δ|B-L|=0$, and $2.91\times10^{-5}$ for the decays $D^{+(-)}\to n(\bar{n})e^{+(-)}$ with $Δ|B-L|=2$ , where $Δ|B-L|$ denotes the change in the difference between baryon and lepton numbers.

preprint2022arXiv

Search for invisible decays of the $Λ$ baryon

A search for invisible decays of the $Λ$ baryon is carried out in the process $J/ψ\toΛ\barΛ$ based on $(1.0087\pm0.0044)\times10^{10}$ $J/ψ$ events collected with the BESIII detector located at the BEPCII storage ring. No signals are found for the invisible decays of $Λ$ baryon, and the upper limit of the branching fraction is determined to be $7.4 \times 10^{-5}$ at the 90% confidence level. This is the first search for invisible decays of baryons; such searches will play an important role in constraining dark sector models related to the baryon asymmetry.

preprint2022arXiv

Search for new hadronic decays of $h_{c}$ and observation of $h_{c}\to p\bar{p}η$

A search for the hadronic decays of the $h_{c}$ meson to the final states $p\bar{p}π^{+}π^{-}π^{0}$, $p\bar{p}η$, and $p\bar{p}π^0$ via the process $ψ(3686) \to π^{0}{h_c}$ is performed using $(4.48\pm0.03)\times10^{8}$ $ψ(3686)$ events collected with the BESIII detector. The decay channel $h_{c}\to p\bar{p}η$ is observed for the first time with a significance greater than $5σ$ and a branching fraction of $\left( {6.41 \pm 1.74 \pm 0.53 \pm 1.00} \right) \times {10^{ -4}}$, where the uncertainties are statistical, systematic, and that from the branching fraction of $ψ(3686)\toπ^{0}h_{c}$. Strong evidence for the decay ${h_c} \to p\bar{p}{π^+}{π^-}{π^0}$ is found with a significance of $4.9σ$ and a branching fraction of $\left( {3.84 \pm 0.83 \pm0.69} \pm 0.58 \right) \times {10^{ - 3}}$. The significances include systematic uncertainties. No clear signal of the decay $h_c\to p\bar{p}π^{0}$ is found, and an upper limit of $6.59\times 10^{-4}$ on its branching fraction is set at the 90% confidence level.

preprint2022arXiv

Search for the decay $D^{0} \to π^{0} ν\barν$

We present the first experimental search for the rare charm decay $D^{0} \to π^{0} ν\barν$. It is based on an $e^+e^-$ collision sample consisting of $10.6\times10^{6}$ pairs of $D^0\bar{D}^0$ mesons collected by the BESIII detector at $\sqrt{s}$=3.773 GeV, corresponding to an integrated luminosity of 2.93~fb$^{-1}$. A data-driven method is used to ensure the reliability of the background modeling. No significant $D^{0} \to π^{0} ν\barν$ signal is observed in data and an upper limit of the branching fraction is set to be $2.1\times 10^{-4}$ at the 90$\%$ confidence level. This is the first experimental constraint on charmed-hadron decays into dineutrino final states.

preprint2022arXiv

Search for the decay $h_c\rightarrowπ^0J/ψ$

A search for the decay $h_c\rightarrowπ^0J/ψ$ is performed using a sample of $h_c$ produced in the reaction $e^+e^-\rightarrowπ^+π^-h_c$. The data samples were collected with the BESIII detector at center-of-mass energies between 4.189 and 4.437 GeV, corresponding to a total integrated luminosity of 11 fb$^{-1}$. No significant signal is observed. Upper limits on the branching ratio $\mathcal{B}(h_c\rightarrowπ^0J/ψ)/\mathcal{B}(h_c\rightarrowγη_c\rightarrowγK^+K^-π^0)$ and on the branching fraction $\mathcal{B}(h_c\rightarrowπ^0J/ψ)$ are determined to be $7.5\times10^{-2}$ and $4.7\times10^{-4}$ at $90\%$ confidence level, respectively. The latter is derived from the former using the measured branching fraction of the normalization channel. This is the first determination of the upper limit of the decay $h_c\rightarrowπ^0J/ψ$.

preprint2022arXiv

Skin Lesion Recognition with Class-Hierarchy Regularized Hyperbolic Embeddings

In practice, many medical datasets have an underlying taxonomy defined over the disease label space. However, existing classification algorithms for medical diagnoses often assume semantically independent labels. In this study, we aim to leverage class hierarchy with deep learning algorithms for more accurate and reliable skin lesion recognition. We propose a hyperbolic network to learn image embeddings and class prototypes jointly. The hyperbola provably provides a space for modeling hierarchical relations better than Euclidean geometry. Meanwhile, we restrict the distribution of hyperbolic prototypes with a distance matrix that is encoded from the class hierarchy. Accordingly, the learned prototypes preserve the semantic class relations in the embedding space and we can predict the label of an image by assigning its feature to the nearest hyperbolic class prototype. We use an in-house skin lesion dataset which consists of around 230k dermoscopic images on 65 skin diseases to verify our method. Extensive experiments provide evidence that our model can achieve higher accuracy with less severe classification errors than models without considering class relations.

preprint2022arXiv

SP-ViT: Learning 2D Spatial Priors for Vision Transformers

Recently, transformers have shown great potential in image classification and established state-of-the-art results on the ImageNet benchmark. However, compared to CNNs, transformers converge slowly and are prone to overfitting in low-data regimes due to the lack of spatial inductive biases. Such spatial inductive biases can be especially beneficial since the 2D structure of an input image is not well preserved in transformers. In this work, we present Spatial Prior-enhanced Self-Attention (SP-SA), a novel variant of vanilla Self-Attention (SA) tailored for vision transformers. Spatial Priors (SPs) are our proposed family of inductive biases that highlight certain groups of spatial relations. Unlike convolutional inductive biases, which are forced to focus exclusively on hard-coded local regions, our proposed SPs are learned by the model itself and take a variety of spatial relations into account. Specifically, the attention score is calculated with emphasis on certain kinds of spatial relations at each head, and such learned spatial foci can be complementary to each other. Based on SP-SA we propose the SP-ViT family, which consistently outperforms other ViT models with similar GFlops or parameters. Our largest model SP-ViT-L achieves a record-breaking 86.3% Top-1 accuracy with a reduction in the number of parameters by almost 50% compared to previous state-of-the-art model (150M for SP-ViT-L vs 271M for CaiT-M-36) among all ImageNet-1K models trained on 224x224 and fine-tuned on 384x384 resolution w/o extra data.

preprint2022arXiv

Spatiotemporal Self-attention Modeling with Temporal Patch Shift for Action Recognition

Transformer-based methods have recently achieved great advancement on 2D image-based vision tasks. For 3D video-based tasks such as action recognition, however, directly applying spatiotemporal transformers on video data will bring heavy computation and memory burdens due to the largely increased number of patches and the quadratic complexity of self-attention computation. How to efficiently and effectively model the 3D self-attention of video data has been a great challenge for transformers. In this paper, we propose a Temporal Patch Shift (TPS) method for efficient 3D self-attention modeling in transformers for video-based action recognition. TPS shifts part of patches with a specific mosaic pattern in the temporal dimension, thus converting a vanilla spatial self-attention operation to a spatiotemporal one with little additional cost. As a result, we can compute 3D self-attention using nearly the same computation and memory cost as 2D self-attention. TPS is a plug-and-play module and can be inserted into existing 2D transformer models to enhance spatiotemporal feature learning. The proposed method achieves competitive performance with state-of-the-arts on Something-something V1 & V2, Diving-48, and Kinetics400 while being much more efficient on computation and memory cost. The source code of TPS can be found at https://github.com/MartinXM/TPS.

preprint2022arXiv

SwinFuse: A Residual Swin Transformer Fusion Network for Infrared and Visible Images

The existing deep learning fusion methods mainly concentrate on the convolutional neural networks, and few attempts are made with transformer. Meanwhile, the convolutional operation is a content-independent interaction between the image and convolution kernel, which may lose some important contexts and further limit fusion performance. Towards this end, we present a simple and strong fusion baseline for infrared and visible images, namely\textit{ Residual Swin Transformer Fusion Network}, termed as SwinFuse. Our SwinFuse includes three parts: the global feature extraction, fusion layer and feature reconstruction. In particular, we build a fully attentional feature encoding backbone to model the long-range dependency, which is a pure transformer network and has a stronger representation ability compared with the convolutional neural networks. Moreover, we design a novel feature fusion strategy based on $L_{1}$-norm for sequence matrices, and measure the corresponding activity levels from row and column vector dimensions, which can well retain competitive infrared brightness and distinct visible details. Finally, we testify our SwinFuse with nine state-of-the-art traditional and deep learning methods on three different datasets through subjective observations and objective comparisons, and the experimental results manifest that the proposed SwinFuse obtains surprising fusion performance with strong generalization ability and competitive computational efficiency. The code will be available at https://github.com/Zhishe-Wang/SwinFuse.

preprint2022arXiv

The $ω^3$ scaling of the vibrational density of states in quasi-2D nanoconfined solids

Atomic vibrations play a vital role in the functions of various physical, chemical, and biological systems. The vibrational properties and the specific heat of crystalline bulk materials are well described by Debye theory, which successfully predicts the quadratic $ω^{2}$ low-frequency scaling of the vibrational density of states (VDOS) in bulk ordered solids from few fundamental assumptions. However, the analogous framework for nanoconfined materials with fewer degrees of freedom has been far less well explored. Using inelastic neutron scattering, we characterize the VDOS of amorphous ice confined to a thickness of $\approx 1$ nm inside graphene oxide membranes and we observe a crossover from the Debye $ω^2$ scaling to an anomalous $ω^3$ behaviour upon reducing the confinement size $L$. Additionally, using molecular dynamics simulations, we confirm the experimental findings and also prove that such a scaling of the VDOS appears in both crystalline and amorphous solids under slab-confinement. We theoretically demonstrate that this low-frequency $ω^3$ law results from the geometric constraints on the momentum phase space induced by confinement along one spatial direction. Finally, we predict that the Debye scaling reappears at a characteristic frequency $ω_\times= v L/2π$, with $v$ the speed of sound of the material, and we confirm this quantitative estimate with simulations. This new physical phenomenon, revealed by combining theoretical, experimental and simulations results, is relevant to a myriad of systems both in synthetic and biological contexts and it could impact various technological applications for systems under confinement such as nano-devices or thin films.

preprint2022arXiv

The Benefit of Hindsight: Tracing Edge-Cases in Distributed Systems

Today's distributed tracing frameworks are ill-equipped to troubleshoot rare edge-case requests. The crux of the problem is a trade-off between specificity and overhead. On the one hand, frameworks can indiscriminately select requests to trace when they enter the system (head sampling), but this is unlikely to capture a relevant edge-case trace because the framework cannot know which requests will be problematic until after-the-fact. On the other hand, frameworks can trace everything and later keep only the interesting edge-case traces (tail sampling), but this has high overheads on the traced application and enormous data ingestion costs. In this paper we circumvent this trade-off for any edge-case with symptoms that can be programmatically detected, such as high tail latency, errors, and bottlenecked queues. We propose a lightweight and always-on distributed tracing system, Hindsight, which implements a retroactive sampling abstraction: instead of eagerly ingesting and processing traces, Hindsight lazily retrieves trace data only after symptoms of a problem are detected. Hindsight is analogous to a car dash-cam that, upon detecting a sudden jolt in momentum, persists the last hour of footage. Developers using Hindsight receive the exact edge-case traces they desire without undue overhead or dependence on luck. Our evaluation shows that Hindsight scales to millions of requests per second, adds nanosecond-level overhead to generate trace data, handles GB/s of data per node, transparently integrates with existing distributed tracing systems, and successfully persists full, detailed traces in real-world use cases when edge-case problems are detected.

preprint2022arXiv

The Blow-up Analysis on $\mathbf{B}_2^{(1)}$ Affine Toda system: Local mass and Affine Weyl group

It has been established that the local mass of blow-up solutions to Toda systems associated with the simple Lie algebras $\mathbf{A}_n,~\mathbf{B}_n,~\mathbf{C}_n$ and $\mathbf{G}_2$ can be represented by a finite Weyl group. In particular, at each blow-up point, after a sequence of bubbling steps (via scaling) is performed, the transformation of the local mass at each step corresponds to the action of an element in the Weyl group. In this article, we present the results in the same spirit for the affine $\mathbf{B}_2^{(1)}$ Toda system with singularities. Compared with the Toda system with simple Lie algebras, the computation of local masses is more challenging due to the infinite number of elements of the {affine Weyl group of type $\mathbf{B}_{2}^{(1)}$}. In order to give an explicit expression for the local mass formula we introduce two free integers and write down all the possibilities into 8 types. This shows a striking difference to previous results on Toda systems with simple Lie algebras. The main result of this article seems to provide the first major advance in understanding the relation between the blow-up analysis of affine Toda system and the {affine Weyl group} of the associated Lie algebras.

preprint2022arXiv

The Nonequilibrium Mechanism of Noise Enhancer synergizing with Activator in HIV Latency Reactivation

Noise-modulating chemicals can synergize with transcriptional activators in reactivating latent HIV to eliminate latent HIV reservoirs. To understand the underlying biomolecular mechanism, we investigate a previous two-gene-state model and identify two necessary conditions for the synergy: an assumption of inhibition effect of transcription activators on noise enhancers; and frequent transitions to the gene non-transcription-permissive state. We then develop a loop-four-gene-state model with Tat transcription/translation and find that drug synergy is mainly determined by the magnitude and direction of energy input into the genetic regulatory kinetics of the HIV promoter. The inhibition effect of transcription activators is actually a phenomenon of energy dissipation in the nonequilibrium gene transition system. Overall, the loop-four-state model demonstrates that energy dissipation plays a crucial role in HIV latency reactivation, which might be useful for improving drug effects and identifying other synergies on lentivirus latency reactivation.

preprint2022arXiv

Theoretical Study of Elastic Far-Field Decay from Dislocations in Multilattices

We precisely and rigorously characterise the decay of elastic fields generated by dislocations in crystalline materials, focusing specifically on the role of multilattices. Concretely, we establish that the elastic field generated by a dislocation in a multilattice can be decomposed into a continuum field predicted by a linearised Cauchy-Born elasticity theory, and a discrete and nonlinear core corrector representing the defect core. We demonstrate both analytically and numerically the consequences of this result for cell size effects in numerical simulations.

preprint2022arXiv

Towards Robust 2D Convolution for Reliable Visual Recognition

2D convolution (Conv2d), which is responsible for extracting features from the input image, is one of the key modules of a convolutional neural network (CNN). However, Conv2d is vulnerable to image corruptions and adversarial samples. It is an important yet rarely investigated problem that whether we can design a more robust alternative of Conv2d for more reliable feature extraction. In this paper, inspired by the recently developed learnable sparse transform that learns to convert the CNN features into a compact and sparse latent space, we design a novel building block, denoted by RConv-MK, to strengthen the robustness of extracted convolutional features. Our method leverages a set of learnable kernels of different sizes to extract features at different frequencies and employs a normalized soft thresholding operator to adaptively remove noises and trivial features at different corruption levels. Extensive experiments on clean images, corrupted images as well as adversarial samples validate the effectiveness of the proposed robust module for reliable visual recognition. The source codes are enclosed in the submission.

preprint2022arXiv

Unfolded Deep Kernel Estimation for Blind Image Super-resolution

Blind image super-resolution (BISR) aims to reconstruct a high-resolution image from its low-resolution counterpart degraded by unknown blur kernel and noise. Many deep neural network based methods have been proposed to tackle this challenging problem without considering the image degradation model. However, they largely rely on the training sets and often fail to handle images with unseen blur kernels during inference. Deep unfolding methods have also been proposed to perform BISR by utilizing the degradation model. Nonetheless, the existing deep unfolding methods cannot explicitly solve the data term of the unfolding objective function, limiting their capability in blur kernel estimation. In this work, we propose a novel unfolded deep kernel estimation (UDKE) method, which, for the first time to our best knowledge, explicitly solves the data term with high efficiency. The UDKE based BISR method can jointly learn image and kernel priors in an end-to-end manner, and it can effectively exploit the information in both training data and image degradation model. Experiments on benchmark datasets and real-world data demonstrate that the proposed UDKE method could well predict complex unseen non-Gaussian blur kernels in inference, achieving significantly better BISR performance than state-of-the-art. The source code of UDKE is available at: https://github.com/natezhenghy/UDKE.

preprint2022arXiv

Universal Domain Adaptive Object Detector

Universal domain adaptive object detection (UniDAOD)is more challenging than domain adaptive object detection (DAOD) since the label space of the source domain may not be the same as that of the target and the scale of objects in the universal scenarios can vary dramatically (i.e, category shift and scale shift). To this end, we propose US-DAF, namely Universal Scale-Aware Domain Adaptive Faster RCNN with Multi-Label Learning, to reduce the negative transfer effect during training while maximizing transferability as well as discriminability in both domains under a variety of scales. Specifically, our method is implemented by two modules: 1) We facilitate the feature alignment of common classes and suppress the interference of private classes by designing a Filter Mechanism module to overcome the negative transfer caused by category shift. 2) We fill the blank of scale-aware adaptation in object detection by introducing a new Multi-Label Scale-Aware Adapter to perform individual alignment between the corresponding scale for two domains. Experiments show that US-DAF achieves state-of-the-art results on three scenarios (i.e, Open-Set, Partial-Set, and Closed-Set) and yields 7.1% and 5.9% relative improvement on benchmark datasets Clipart1k and Watercolor in particular.

preprint2022arXiv

Vanishing Estimates for Liouville equation with quantized singularities

In this article we continue with the research initiated in our previous work on singular Liouville equations with quantized singularity. The main goal of this article is to prove that as long as the bubbling solutions violate the spherical Harnack inequality near a singular source, the first derivatives of coefficient functions must tend to zero.

preprint2022arXiv

Vision-Language Intelligence: Tasks, Representation Learning, and Large Models

This paper presents a comprehensive survey of vision-language (VL) intelligence from the perspective of time. This survey is inspired by the remarkable progress in both computer vision and natural language processing, and recent trends shifting from single modality processing to multiple modality comprehension. We summarize the development in this field into three time periods, namely task-specific methods, vision-language pre-training (VLP) methods, and larger models empowered by large-scale weakly-labeled data. We first take some common VL tasks as examples to introduce the development of task-specific methods. Then we focus on VLP methods and comprehensively review key components of the model structures and training methods. After that, we show how recent work utilizes large-scale raw image-text data to learn language-aligned visual representations that generalize better on zero or few shot learning tasks. Finally, we discuss some potential future trends towards modality cooperation, unified representation, and knowledge incorporation. We believe that this review will be of help for researchers and practitioners of AI and ML, especially those interested in computer vision and natural language processing.

preprint2022arXiv

Voxel Set Transformer: A Set-to-Set Approach to 3D Object Detection from Point Clouds

Transformer has demonstrated promising performance in many 2D vision tasks. However, it is cumbersome to compute the self-attention on large-scale point cloud data because point cloud is a long sequence and unevenly distributed in 3D space. To solve this issue, existing methods usually compute self-attention locally by grouping the points into clusters of the same size, or perform convolutional self-attention on a discretized representation. However, the former results in stochastic point dropout, while the latter typically has narrow attention fields. In this paper, we propose a novel voxel-based architecture, namely Voxel Set Transformer (VoxSeT), to detect 3D objects from point clouds by means of set-to-set translation. VoxSeT is built upon a voxel-based set attention (VSA) module, which reduces the self-attention in each voxel by two cross-attentions and models features in a hidden space induced by a group of latent codes. With the VSA module, VoxSeT can manage voxelized point clusters with arbitrary size in a wide range, and process them in parallel with linear complexity. The proposed VoxSeT integrates the high performance of transformer with the efficiency of voxel-based model, which can be used as a good alternative to the convolutional and point-based backbones. VoxSeT reports competitive results on the KITTI and Waymo detection benchmarks. The source codes can be found at \url{https://github.com/skyhehe123/VoxSeT}.

preprint2022arXiv

Winograd Convolution: A Perspective from Fault Tolerance

Winograd convolution is originally proposed to reduce the computing overhead by converting multiplication in neural network (NN) with addition via linear transformation. Other than the computing efficiency, we observe its great potential in improving NN fault tolerance and evaluate its fault tolerance comprehensively for the first time. Then, we explore the use of fault tolerance of winograd convolution for either fault-tolerant or energy-efficient NN processing. According to our experiments, winograd convolution can be utilized to reduce fault-tolerant design overhead by 27.49\% or energy consumption by 7.19\% without any accuracy loss compared to that without being aware of the fault tolerance

preprint2021arXiv

A crystalline incarnation of Berthelot's conjecture and Künneth formula for isocrystals

Berthelot's conjecture predicts that under a proper and smooth morphism of schemes in characteristic $p$, the higher direct images of an overconvergent $F$-isocrystal are overconvergent $F$-isocrystals. In this paper we prove that this is true for crystals up to isogeny. As an application we prove a Künneth formula for the crystalline fundamental group.

preprint2021arXiv

A Mean Field Game Analysis of Consensus Protocol Design

A decentralized blockchain is a distributed ledger that is often used as a platform for exchanging goods and services. This ledger is maintained by a network of nodes that obeys a set of rules, called a consensus protocol, which helps to resolve inconsistencies among local copies of a blockchain. In this paper, we build a mathematical framework for the consensus protocol designer, specifying (a) the measurement of a resource which nodes strategically invest in and compete for to win the right to build new blocks in the blockchain; and (b) a payoff function for such efforts. Thus, the equilibrium of an associated stochastic differential game can be implemented by selecting nodes in proportion to this specified resource and penalizing dishonest nodes by its loss. This associated, induced game can be further analyzed using mean field games. The problem can be broken down into two coupled PDEs, where an individual node's optimal control path is solved using a Hamilton-Jacobi-Bellman equation, and where the evolution of states distribution is characterized by a Fokker-Planck equation. We develop numerical methods to compute the mean field equilibrium for both steady states at the infinite time horizon and evolutionary dynamics. As an example, we show how the mean field equilibrium can be applied to the Bitcoin blockchain mechanism design. We demonstrate that a blockchain can be viewed as a mechanism that operates in a decentralized setup and propagates properties of the mean field equilibrium over time, such as the underlying security of the blockchain.

preprint2021arXiv

Accurate Mode-Coupling Characterization of Low-Crosstalk Ring-Core Fibers using Integral Calculation based Swept-Wavelength Interferometry Measurement

In this paper, to accurately characterize the low inter-mode coupling of the weakly-coupled few mode fibers (FMFs), we propose a modified inter-mode coupling characterization method based on swept-wavelength interferometry measurement, in which an integral calculation approach is used to eliminate significant sources of error that may lead to underestimation of the power coupling coefficient. Using the proposed characterization method, a low-crosstalk ring-core fiber (RCF) with low mode dependent loss (MDL) and with single span length up to 100 km is experimentally measured to have low power coupling coefficients between high-order orbital angular momentum (OAM) mode groups of below -30 dB/km over C band. The measured low coupling coefficients based on the proposed method are verified by the direct system power measurements, proving the feasibility and reliability of the proposed inter-mode coupling characterization method.

preprint2021arXiv

Computing solution landscape of nonlinear space-fractional problems via fast approximation algorithm

The nonlinear space-fractional problems often allow multiple stationary solutions, which can be much more complicated than the corresponding integer-order problems. In this paper, we systematically compute the solution landscapes of nonlinear constant/variable-order space-fractional problems. A fast approximation algorithm is developed to deal with the variable-order spectral fractional Laplacian by approximating the variable-indexing Fourier modes, and then combined with saddle dynamics to construct the solution landscape of variable-order space-fractional phase field model. Numerical experiments are performed to substantiate the accuracy and efficiency of fast approximation algorithm and elucidate essential features of the stationary solutions of space-fractional phase field model. Furthermore, we demonstrate that the solution landscapes of spectral fractional Laplacian problems can be reconfigured by varying the diffusion coefficients in the corresponding integer-order problems.

preprint2021arXiv

Constructing Evacuation Evolution Patterns and Decisions Using Mobile Device Location Data: A Case Study of Hurricane Irma

Understanding individuals' behavior during hurricane evacuation is of paramount importance for local, state, and government agencies hoping to be prepared for natural disasters. Complexities involved with human decision-making procedures and lack of data for such disasters are the main reasons that make hurricane evacuation studies challenging. In this paper, we utilized a large mobile phone Location-Based Services (LBS) data to construct the evacuation pattern during the landfall of Hurricane Irma. By employing our proposed framework on more than 11 billion mobile phone location sightings, we were able to capture the evacuation decision of 807,623 smartphone users who were living within the state of Florida. We studied users' evacuation decisions, departure and reentry date distribution, and destination choice. In addition to these decisions, we empirically examined the influence of evacuation order and low-lying residential areas on individuals' evacuation decisions. Our analysis revealed that 57.92% of people living in mandatory evacuation zones evacuated their residences while this ratio was 32.98% and 33.68% for people living in areas with no evacuation order and voluntary evacuation order, respectively. Moreover, our analysis revealed the importance of the individuals' mobility behavior in modeling the evacuation decision choice. Historical mobility behavior information such as number of trips taken by each individual and the spatial area covered by individuals' location trajectory estimated significant in our choice model and improve the overall accuracy of the model significantly.

preprint2021arXiv

Cross section measurements of the $e^+e^-\to D^{+}D^{-}$ and $e^+e^-\to D^{*+}D^{-}$ processes at center-of-mass energies from 4.085 to 4.600 GeV

The Born cross sections of the $e^+e^-\to D^{*+}D^{*-}$ and $e^+e^-\to D^{*+}D^{-}$ processes are measured using $e^+e^-$ collision data collected with the BESIII experiment at center-of-mass energies from 4.085 to 4.600 GeV, corresponding to an integrated luminosity of $15.7~{\rm fb}^{-1}$. The results are consistent with and more precise than the previous measurements by the Belle, Babar and CLEO collaborations. The measurements are essential for understanding the nature of vector charmonium and charmonium-like states.

preprint2021arXiv

Cross sections for the reactions $e^+e^-\rightarrow K^+K^-π^+π^-(π^0)$, $K^+K^-K^+K^-(π^0)$, $π^+π^-π^+π^-(π^0)$, $p\bar{p}π^+π^-(π^0)$ in the energy region between 3.773 and 4.600 GeV

Using the data samples collected in the energy range from 3.773 to 4.600 GeV with the BESIII detector at the BEPCII collider, we measure the dressed cross sections as a function of center-of-mass energy for $e^+e^-\rightarrow K^+K^-π^+π^-(π^0)$, $K^+K^-K^+K^-(π^0)$, $π^+π^-π^+π^-(π^0)$, and $p\bar{p}π^+π^-(π^0)$. The cross sections for $e^+e^-\rightarrow K^+K^-K^+K^-π^0$, $p\bar{p}π^+π^-(π^0)$ are the first measurements. Cross sections for the other five channels are much more precise than previous results in this energy region. We also search for charmonium and charmonium-like resonances, such as the $Y(4230)$, decaying into the same final states. We find evidence of the $ψ(4040)$ decaying to $π^+π^-π^+π^-π^0$ with a statistical significance of $3.6σ$. Upper limits are provided for other decays since no clear signals are observed.

preprint2021arXiv

Defect patterns of two-dimensional nematic liquid crystals in confinement

A two-dimensional or quasi-two-dimensional nematic liquid crystal refers to a surface confined system. When such a system is further confined by external line boundaries or excluded from internal line boundaries, the nematic directors form a deformed texture that may display defect points or defect lines, for which winding numbers can be clearly defined. Here, a particular attention is paid to the case when the liquid crystal molecules prefer to form a boundary nematic texture in parallel to the wall surface (i.e., following the homogeneous boundary condition). A general theory, based on geometric argument, is presented for the relationship between the sum of all winding numbers in the system (the total winding number) and the type of confinement angles and curved segments. The conclusion is validated by comparing the theoretical defect rule with existing nematic textures observed experimentally and theoretically in recent years.

preprint2021arXiv

Dual MINE-based Neural Secure Communications under Gaussian Wiretap Channel

Recently, some researches are devoted to the topic of end-to-end learning a physical layer secure communication system based on autoencoder under Gaussian wiretap channel. However, in those works, the reliability and security of the encoder model were learned through necessary decoding outputs of not only legitimate receiver but also the eavesdropper. In fact, the assumption of known eavesdropper's decoder or its output is not practical. To address this issue, in this paper we propose a dual mutual information neural estimation (MINE) based neural secure communications model. The security constraints of this method is constructed only with the input and output signal samples of the legal and eavesdropper channels and benefit that training the encoder is completely independent of the decoder. Moreover, since the design of secure coding does not rely on the eavesdropper's decoding results, the security performance would not be affected by the eavesdropper's decoding means. Numerical results show that the performance of our model is guaranteed whether the eavesdropper learns the decoder himself or uses the legal decoder.

preprint2021arXiv

Estimates for Liouville equation with quantized singularities

For Liouville equations with singular sources, the interpretation of the equation and its impact are most significant if the singular sources are quantized: the strength of each Dirac mass is a mutliple of $4π$. However the study of bubbling solutions around a quantized singular source is particularly challenging: near the singular source, the spherical Harnack inequality may not hold and there are multiple local maximums of bubbling solutions all swarming to the singular source. In this article we seek to provide a complete understanding of the blowup picture in this core difficulty and we establish two major types of results: First we prove that not only the first derivatives of coefficient functions tend to zero at the singular source, the second derivatives also have a vanishing estimate. Second we derive pointwise estimates for bubbling solutions to be approximated by global solutions, which have crucial applications in a number of important projects. Since bubbling solutions near a singular source are very commonly observed in geometry and physics, it seems that the estimates in this article can also be applied to many related equations and systems with various backgrounds.

preprint2021arXiv

Generalized Rough Polyharmonic Splines for Multiscale PDEs with Rough Coefficients

In this paper, we demonstrate the construction of generalized Rough Polyhamronic Splines (GRPS) within the Bayesian framework, in particular, for multiscale PDEs with rough coefficients. The optimal coarse basis can be derived automatically by the randomization of the original PDEs with a proper prior distribution and the conditional expectation given partial information on edge or derivative measurements. We prove the (quasi)-optimal localization and approximation properties of the obtained bases, and justify the theoretical results with numerical experiments.

preprint2021arXiv

HiDeNN-PGD: reduced-order hierarchical deep learning neural networks

This paper presents a proper generalized decomposition (PGD) based reduced-order model of hierarchical deep-learning neural networks (HiDeNN). The proposed HiDeNN-PGD method keeps both advantages of HiDeNN and PGD methods. The automatic mesh adaptivity makes the HiDeNN-PGD more accurate than the finite element method (FEM) and conventional PGD, using a fraction of the FEM degrees of freedom. The accuracy and convergence of the method have been studied theoretically and numerically, with a comparison to different methods, including FEM, PGD, HiDeNN and Deep Neural Networks. In addition, we theoretically showed that the PGD converges to FEM at increasing modes, and the PGD error is a direct sum of the FEM error and the mode reduction error. The proposed HiDeNN-PGD performs high accuracy with orders of magnitude fewer degrees of freedom, which shows a high potential to achieve fast computations with a high level of accuracy for large-size engineering problems.

preprint2021arXiv

How Much Communication Resource is Needed to Run a Wireless Blockchain Network?

Blockchain is built on a peer-to-peer network that relies on frequent communications among the distributively located nodes. In particular, the consensus mechanisms (CMs), which play a pivotal role in blockchain, are communication resource-demanding and largely determines blockchain security bound and other key performance metrics such as transaction throughput, latency and scalability. Most blockchain systems are designed in a stable wired communication network running in advanced devices under the assumption of sufficient communication resource provision. However, it is envisioned that the majority of the blockchain node peers will be connected through the wireless network in the future. Constrained by the highly dynamic wireless channel and scarce frequency spectrum, communication can significantly affect blockchain's key performance metrics. Hence, in this paper, we present wireless blockchain networks (WBN) under various commonly used CMs and we answer the question of how much communication resource is needed to run such a network. We first present the role of communication in the four stages of the blockchain procedure. We then discuss the relationship between the communication resource provision and the WBNs performance, for three of the most used blockchain CMs namely, Proof-of-Work (PoW), practical Byzantine Fault Tolerant (PBFT) and Raft. Finally, we provide analytical and simulated results to show the impact of the communication resource provision on blockchain performance.

preprint2021arXiv

Iterated numerical homogenization for multi-scale elliptic equations with monotone nonlinearity

Nonlinear multi-scale problems are ubiquitous in materials science and biology. Complicated interactions between nonlinearities and (nonseparable) multiple scales pose a major challenge for analysis and simulation. In this paper, we study the numerical homogenization for multi-scale elliptic PDEs with monotone nonlinearity, in particular the Leray-Lions problem (a prototypical example is the p-Laplacian equation), where the nonlinearity cannot be parameterized with low dimensional parameters, and the linearization error is non-negligible. We develop the iterated numerical homogenization scheme by combining numerical homogenization methods for linear equations, and the so-called "quasi-norm" based iterative approach for monotone nonlinear equation. We propose a residual regularized nonlinear iterative method, and in addition, develop the sparse updating method for the efficient update of coarse spaces. A number of numerical results are presented to complement the analysis and valid the numerical method.

preprint2021arXiv

Large-Scale Intelligent Microservices

Deploying Machine Learning (ML) algorithms within databases is a challenge due to the varied computational footprints of modern ML algorithms and the myriad of database technologies each with its own restrictive syntax. We introduce an Apache Spark-based micro-service orchestration framework that extends database operations to include web service primitives. Our system can orchestrate web services across hundreds of machines and takes full advantage of cluster, thread, and asynchronous parallelism. Using this framework, we provide large scale clients for intelligent services such as speech, vision, search, anomaly detection, and text analysis. This allows users to integrate ready-to-use intelligence into any datastore with an Apache Spark connector. To eliminate the majority of overhead from network communication, we also introduce a low-latency containerized version of our architecture. Finally, we demonstrate that the services we investigate are competitive on a variety of benchmarks, and present two applications of this framework to create intelligent search engines, and real-time auto race analytics systems.

preprint2021arXiv

Measurement of Branching Fractions of $J/ψ$ and $ψ(3686)$ decays to $Σ^{+}$ and $\overlineΣ^-$

Using $1310.6\times10^{6}$ $J/ψ$ and $448.1\times10^{6}$ $ψ(3686)$ events collected with the BESIII detector, the branching fractions of $J/ψ$ and $ψ(3686)$ decays to $Σ^{+}\overlineΣ^{-}$ are measured to be $(10.61 \pm 0.04 \pm 0.36) \times 10^{-4}$ and $(2.52 \pm 0.04 \pm 0.09) \times 10^{-4}$, respectively. In addition, the ratio of $\mathcal{B}(ψ(3686) \rightarrow Σ^{+}\overlineΣ^{-})/\mathcal{B}(J/ψ\rightarrow Σ^{+}\overlineΣ^{-})$ is determined to be $(23.8 \pm 1.1)\%$ which violates the "$12\%$ rule".

preprint2021arXiv

Measurement of cross-section for $e^+e^-\toΞ^-\barΞ^+$ near threshold at BESIII

The Born cross-sections and effective form factors for process $e^+e^-\toΞ^-\barΞ^+$ are measured at eight center-of-mass energies between 2.644 and 3.080 GeV, using a total integrated luminosity of 363.9 pb$^{-1}$ $e^+e^-$ collision data collected with the BESIII detector at BEPCII. After performing a fit to the Born cross-section of $e^+e^-\toΞ^-\barΞ^+$, no significant threshold effect is observed.

preprint2021arXiv

Measurement of the $e^{+}e^{-}\toΣ^{0}\barΣ^{0}$ cross sections at center-of-mass energies from $2.3864$ to $3.0200$ GeV

The Born cross sections of $e^{+}e^{-}\to Σ^{0}\barΣ^{0}$ are measured at center-of-mass energies from $2.3864$ to $3.0200$ GeV using data samples with an integrated luminosity of $328.5$ pb$^{-1}$ collected with the BESIII detector operating at the BEPCII collider. The analysis makes use of a novel reconstruction method for energies near production threshold, while a single-tag method is employed at other center-of-mass energies. The measured cross sections are consistent with earlier results from BaBar, with a substantially improved precision. The cross-section lineshape can be well described by a perturbative QCD-driven energy function. In addition, the effective form factors of the $Σ^{0}$ baryon are determined. The results provide precise experimental input for testing various theoretical predictions.

preprint2021arXiv

Measurements of $e^+e^-\rightarrow η_{\rm c}π^+ π^-π^0$, $η_{\rm c}π^+ π^-$ and $η_{\rm c}π^0γ$ at $\sqrt{s}$ from 4.18 to 4.60\,GeV, and search for a $Z_{\rm c}$ state close to the $D\bar{D}$ threshold decaying to $η_{\rm c}π$ at $\sqrt{s}$ = 4.23 GeV

We study $η_{\rm c}$ production at center-of-mass energies $\sqrt{s}$ from 4.18 to 4.60 GeV in $e^+e^-$ annihilation data collected with the BESIII detector operating at the BEPCII storage ring, corresponding to 7.3 fb$^{-1}$ of integrated luminosity. We measure the cross sections of the three different exclusive reactions $e^+e^-\rightarrow η_{\rm c}π^+ π^-π^0$, $e^+e^- \rightarrow η_{\rm c}π^+ π^-$, and $e^+e^- \rightarrow η_{\rm c}π^0γ$. We find significant $η_{\rm c}$ production in $e^+e^-\rightarrow η_{\rm c}π^+ π^-π^0$ at $\sqrt{s}$ of 4.23 GeV and 4.26 GeV and observe a significant energy-dependent Born cross section that we measure to be consistent with the production via the intermediate $Y(4260)$ resonance. In addition, we perform a search for a charmonium-like $Z_{\rm c}$ state close to the $D\bar{D}$ threshold that decays to $η_{\rm c}π$, involving ground state charmonium, and observe no signal. Corresponding upper limits on the cross section of $η_{\rm c}$ and $Z_{\rm c}$ production are provided, where the yields are not found to be significant.

preprint2021arXiv

Model independent determination of the spin of the $Ω^{-}$ and its polarization alignment in $ψ(3686)\rightarrowΩ^{-}\barΩ^{+}$

We present an analysis of the process $ψ(3686) \to Ω^- \barΩ^+$ ($Ω^-\to K^-Λ$, $\barΩ^+\to K^+\barΛ$, $Λ\to pπ^-$, $\barΛ\to \bar{p}π^+$) based on a data set of $448\times 10^6$ $ψ(3686)$ decays collected with the BESIII detector at the BEPCII electron-positron collider. The helicity amplitudes for the process $ψ(3686) \to Ω^- \barΩ^+$ and the decay parameters of the subsequent decay $Ω^-\to K^-Λ$ $(\barΩ^+\to K^+\barΛ)$ are measured for the first time by a fit to the angular distribution of the complete decay chain. The branching fraction of $ψ(3686) \to Ω^- \barΩ^+$ is measured to be $(5.82\pm 0.12\pm 0.24)\times 10^{-5}$, with an improved precision compared to previous measurements.

preprint2021arXiv

MosAIc: Finding Artistic Connections across Culture with Conditional Image Retrieval

We introduce MosAIc, an interactive web app that allows users to find pairs of semantically related artworks that span different cultures, media, and millennia. To create this application, we introduce Conditional Image Retrieval (CIR) which combines visual similarity search with user supplied filters or "conditions". This technique allows one to find pairs of similar images that span distinct subsets of the image corpus. We provide a generic way to adapt existing image retrieval data-structures to this new domain and provide theoretical bounds on our approach's efficiency. To quantify the performance of CIR systems, we introduce new datasets for evaluating CIR methods and show that CIR performs non-parametric style transfer. Finally, we demonstrate that our CIR data-structures can identify "blind spots" in Generative Adversarial Networks (GAN) where they fail to properly model the true data distribution.

preprint2021arXiv

Observation of $e^{+}e^{-}\rightarrowηψ(2S)$ at center-of-mass energies from 4.236 to 4.600 GeV

Using a total of $5.25~{\rm fb}^{-1}$ of $e^{+}e^{-}$ collision data with center-of-mass energies from 4.236 to 4.600 GeV, we report the first observation of the process $e^{+}e^{-}\to ηψ(2S)$ with a statistical significance of $5σ$. The data sets were collected by the BESIII detector operating at the BEPCII storage ring. We measure the yield of events integrated over center-of-mass energies and also present the energy dependence of the measured cross section.

preprint2021arXiv

On Liouville systems at critical parameters, Part 2: Multiple bubbles

In this paper, we continue to consider the generalized Liouville system: $$ Δ_g u_i+\sum_{j=1}^n a_{ij}ρ_j\left(\frac{h_j e^{u_j}}{\int h_j e^{u_j}}- {1} \right)=0\quad\text{in \,}M,\quad i\in I=\{1,\cdots,n\}, $$ where $(M,g)$ is a Riemann surface $M$ with volume $1$, $h_1,..,h_n$ are positive smooth functions and $ρ_j\in \mathbb R^+$($j\in I$). In previous works Lin-Zhang identified a family of hyper-surfaces $Γ_N$ and proved a priori estimates for $ρ=(ρ_1,..,ρ_n)$ in areas separated by $Γ_N$. Later Lin-Zhang also calculated the leading term of $ρ^k-ρ$ where $ρ\in Γ_1$ is the limit of $ρ^k$ on $Γ_1$ and $ρ^k$ is the parameter of a bubbling sequence. This leading term is particularly important for applications but it is very hard to be identified if $ρ^k$ tends to a higher order hypersurface $Γ_N$ ($N>1$). Over the years numerous attempts have failed but in this article we overcome all the stumbling blocks and completely solve the problem under the most general context: We not only capture the leading terms of $ρ^k-ρ\in Γ_N$, but also reveal new robustness relations of coefficient functions at different blowup points.

preprint2021arXiv

Pluggable Weakly-Supervised Cross-View Learning for Accurate Vehicle Re-Identification

Learning cross-view consistent feature representation is the key for accurate vehicle Re-identification (ReID), since the visual appearance of vehicles changes significantly under different viewpoints. To this end, most existing approaches resort to the supervised cross-view learning using extensive extra viewpoints annotations, which however, is difficult to deploy in real applications due to the expensive labelling cost and the continous viewpoint variation that makes it hard to define discrete viewpoint labels. In this study, we present a pluggable Weakly-supervised Cross-View Learning (WCVL) module for vehicle ReID. Through hallucinating the cross-view samples as the hardest positive counterparts in feature domain, we can learn the consistent feature representation via minimizing the cross-view feature distance based on vehicle IDs only without using any viewpoint annotation. More importantly, the proposed method can be seamlessly plugged into most existing vehicle ReID baselines for cross-view learning without re-training the baselines. To demonstrate its efficacy, we plug the proposed method into a bunch of off-the-shelf baselines and obtain significant performance improvement on four public benchmark datasets, i.e., VeRi-776, VehicleID, VRIC and VRAI.

preprint2021arXiv

Possible Generation Mechanism for Compressional Alfvénic Spikes as Observed by Parker Solar Probe

The solar wind is found by Parker Solar Probe (PSP) to be abundant with Alfvénic velocity spikes and magnetic field kinks. Temperature enhancement is another remarkable feature associated with the Alfvénic spikes. How the prototype of these coincident phenomena is generated intermittently in the source region becomes a hot topic of wide concerns. Here we propose a new model introducing guide-field discontinuity into the interchange magnetic reconnection between open funnels and closed loops with different magnetic helicities. The modified interchange reconnection model not only can accelerate jet flows from the newly opening closed loop but also excite and launch Alfvénic wave pulses along the newly-reconnected and post-reconnected open flux tubes. We find that the modeling results can reproduce the following observational features: (1) Alfvén disturbance is pulsive in time and asymmetric in space; (2) Alfvénic pulse is compressible with temperature enhancement and density variation inside the pulse. We point out that three physical processes co-happening with Alfvén wave propagation can be responsible for the temperature enhancement: (a) convection of heated jet flow plasmas (decrease in density), (b) propagation of compressed slow-mode waves (increase in density), and (c) conduction of heat flux (weak change in density). We also suggest that the radial nonlinear evolution of the Alfvénic pulses should be taken into account to explain the formation of magnetic switchback geometry.

preprint2021arXiv

Registration-based model reduction of parameterized two-dimensional conservation laws

We propose a nonlinear registration-based model reduction procedure for rapid and reliable solution of parameterized two-dimensional steady conservation laws. This class of problems is challenging for model reduction techniques due to the presence of nonlinear terms in the equations and also due to the presence of parameter-dependent discontinuities that cannot be adequately represented through linear approximation spaces. Our approach builds on a general (i.e., independent of the underlying equation) registration procedure for the computation of a mapping $Φ$ that tracks moving features of the solution field and on an hyper-reduced least-squares Petrov-Galerkin reduced-order model for the rapid and reliable computation of the solution coefficients. The contributions of this work are twofold. First, we investigate the application of registration-based methods to two-dimensional hyperbolic systems. Second, we propose a multi-fidelity approach to reduce the offline costs associated with the construction of the parameterized mapping and the reduced-order model. We discuss the application to an inviscid supersonic flow past a parameterized bump, to illustrate the many features of our method and to demonstrate its effectiveness.

preprint2021arXiv

Reinforcement Learning for Flexibility Design Problems

Flexibility design problems are a class of problems that appear in strategic decision-making across industries, where the objective is to design a ($e.g.$, manufacturing) network that affords flexibility and adaptivity. The underlying combinatorial nature and stochastic objectives make flexibility design problems challenging for standard optimization methods. In this paper, we develop a reinforcement learning (RL) framework for flexibility design problems. Specifically, we carefully design mechanisms with noisy exploration and variance reduction to ensure empirical success and show the unique advantage of RL in terms of fast-adaptation. Empirical results show that the RL-based method consistently finds better solutions compared to classical heuristics.

preprint2021arXiv

Resource Allocation for Mixed Numerology NOMA

6G wireless networks will require the flexibility to accommodate an extremely diverse set of service types. This necessitates the use of mixed numerologies to accommodate different quality of service (QoS) requirements. Non-orthogonal multiple access (NOMA) techniques can potentially be used to accommodate users with different numerologies while also gaining the performance benefits associated with NOMA. To achieve the full performance benefits of a mixed numerology NOMA (MN-NOMA) system, resource allocation among the users is paramount. However, the coexistence of mixed numerologies changes the nature of the interference that each user experiences. This means that techniques used in single-numerology NOMA (SN-NOMA) are no longer sufficient. In light of this, we approach the problem of optimizing subcarrier and power allocation for maximizing the spectral efficiency of MN-NOMA while considering a minimum rate constraint for each user. In this letter, we propose a two-stage sub-optimal approach to solve the problem. We present numerical results which show the superiority of our proposed method over existing benchmark schemes in both spectral efficiency and fairness.

preprint2021arXiv

Room temperature ferromagnetism of monolayer chromium telluride with perpendicular magnetic anisotropy

The realization of long-range magnetic ordering in two-dimensional (2D) systems can potentially revolutionize next-generation information technology. Here, we report the successful fabrication of crystalline Cr3Te4 monolayers with room temperature ferromagnetism. Using molecular beam epitaxy, the growth of 2D Cr3Te4 films with monolayer thickness is demonstrated at low substrate temperatures (~100C), compatible with Si CMOS technology. X-ray magnetic circular dichroism measurements reveal a Curie temperature (Tc) of ~344 K for the Cr3Te4 monolayer with an out-of-plane magnetic easy axis, which decreases to ~240 K for the thicker film (~ 7 nm) with an in-plane easy axis. The enhancement of ferromagnetic coupling and the magnetic anisotropy transition is ascribed to interfacial effects, in particular the orbital overlap at the monolayer Cr3Te4/graphite interface, supported by density-functional theory calculations. This work sheds light on the low-temperature scalable growth of 2D nonlayered materials with room temperature ferromagnetism for new magnetic and spintronic devices.

preprint2021arXiv

Search for the $X(2370)$ and observation of $η_{c}\toηηη^\prime$ in $J/ψ\toγηηη^{\prime}$

Using a sample of $1.31\times10^{9} ~J/ψ$ events collected with the BESIII detector, we perform a study of $J/ψ\toγηηη^{\prime}$ to search for the $X(2370)$ and $η_{c}$ in the $ηηη^{\prime}$ invariant mass distribution. No significant signal for the $X(2370)$ is observed, and we set an upper limit for the product branching fraction of ${\cal B}(J/ψ\toγX(2370)\cdot{\cal B}(X(2370)\toηηη^{\prime}) < 9.2\times10^{-6}$ at the 90% confidence level. A clear $η_{c}$ signal is observed for the first time, yielding a product branching fraction of ${\cal B}(J/ψ\to γη_{c})\cdot{\cal B}(η_{c}\to ηηη^{\prime}) = (4.86\pm0.62~({\rm stat.})\pm0.45~({\rm sys.}))\times10^{-5}$.

preprint2021arXiv

Self-supervised Pre-training with Hard Examples Improves Visual Representations

Self-supervised pre-training (SSP) employs random image transformations to generate training data for visual representation learning. In this paper, we first present a modeling framework that unifies existing SSP methods as learning to predict pseudo-labels. Then, we propose new data augmentation methods of generating training examples whose pseudo-labels are harder to predict than those generated via random image transformations. Specifically, we use adversarial training and CutMix to create hard examples (HEXA) to be used as augmented views for MoCo-v2 and DeepCluster-v2, leading to two variants HEXA_{MoCo} and HEXA_{DCluster}, respectively. In our experiments, we pre-train models on ImageNet and evaluate them on multiple public benchmarks. Our evaluation shows that the two new algorithm variants outperform their original counterparts, and achieve new state-of-the-art on a wide range of tasks where limited task supervision is available for fine-tuning. These results verify that hard examples are instrumental in improving the generalization of the pre-trained models.

preprint2021arXiv

Snapshot Hyperspectral Imaging Based on Weighted High-order Singular Value Regularization

Snapshot hyperspectral imaging can capture the 3D hyperspectral image (HSI) with a single 2D measurement and has attracted increasing attention recently. Recovering the underlying HSI from the compressive measurement is an ill-posed problem and exploiting the image prior is essential for solving this ill-posed problem. However, existing reconstruction methods always start from modeling image prior with the 1D vector or 2D matrix and cannot fully exploit the structurally spectral-spatial nature in 3D HSI, thus leading to a poor fidelity. In this paper, we propose an effective high-order tensor optimization based method to boost the reconstruction fidelity for snapshot hyperspectral imaging. We first build high-order tensors by exploiting the spatial-spectral correlation in HSI. Then, we propose a weight high-order singular value regularization (WHOSVR) based low-rank tensor recovery model to characterize the structure prior of HSI. By integrating the structure prior in WHOSVR with the system imaging process, we develop an optimization framework for HSI reconstruction, which is finally solved via the alternating minimization algorithm. Extensive experiments implemented on two representative systems demonstrate that our method outperforms state-of-the-art methods.

preprint2021arXiv

Superfluid weight and Berezinskii-Kosterlitz-Thouless transition temperature of strained graphene

We obtain the superfluid weight and Berezinskii-Kosterlitz-Thouless (BKT) transition temperature for highly unconventional superconducting states with the coexistence of chiral d-wave superconductivity, charge density waves and pair density waves in the strained graphene. Our results show that the strain-induced flat bands can promote the superconducting transition temperature approximately $50\%$ compared to that of the original doped graphene, which suggests that the flat-band superconductivity is a potential route to get superconductivity with higher critical temperatures. In particular, we obtain the superfluid weight for the pure superconducting pair-density-wave states from which the deduced superconducting transition temperature is shown to be much lower than the gap-opening temperature of the pair density wave, which is helpful to understand the phenomenon of the pseudogap state in high-$T_c$ cuprate superconductors. Finally, we show that the BKT transition temperature versus doping for strained graphene exhibits a dome-like shape and it depends linearly on the spin-spin interaction strength.

preprint2021arXiv

Transition pathways connecting crystals and quasicrystals

Due to structural incommensurability, the emergence of a quasicrystal from a crystalline phase represents a challenge to computational physics. Here the nucleation of quasicrystals is investigated by using an efficient computational method applied to a Landau free-energy functional. Specifically, transition pathways connecting different local minima of the Lifshitz-Petrich model are obtained by using the high-index saddle dynamics. Saddle points on these paths are identified as the critical nuclei of the 6-fold crystals and 12-fold quasicrystals. The results reveal that phase transitions between the crystalline and quasicrystalline phases could follow two possible pathways, corresponding to a one-stage phase transition and a two-stage phase transition involving a metastable lamellar quasicrystalline state, respectively.

preprint2021arXiv

Twist-induced control of near-field heat radiation between magnetic Weyl semimetals

Due to the large anomalous Hall effect, magnetic Weyl semimetals can support nonreciprocal surface plasmon polariton modes in the absence of an external magnetic field. This implies that magnetic Weyl semimetals can find novel application in (thermal) photonics. In this work, we consider the near-field radiative heat transfer between two magnetic Weyl semimetal slabs and show that the heat transfer can be controlled with a relative rotation of the parallel slabs. Thanks to the intrinsic nonreciprocity of the surface modes, this so-called twisting method does not require surface structuring like periodic gratings. The twist-induced control of heat transfer is due to the mismatch of the surface modes from the two slabs with a relative rotation.

preprint2021arXiv

Ultralow complexity long short-term memory network for fiber nonlinearity mitigation in coherent optical communication systems

Fiber Kerr nonlinearity is a fundamental limitation to the achievable capacity of long-distance optical fiber communication. Digital back-propagation (DBP) is a primary methodology to mitigate both linear and nonlinear impairments by solving the inverse-propagating nonlinear Schrödinger equation (NLSE), which requires detailed link information. Recently, the paradigms based on neural network (NN) were proposed to mitigate nonlinear transmission impairments in optical communication systems. However, almost all neural network-based equalization schemes yield high computation complexity, which prevents the practical implementation in commercial transmission systems. In this paper, we propose a center-oriented long short-term memory network (Co-LSTM) incorporating a simplified mode with a recycling mechanism in the equalization operation, which can mitigate fiber nonlinearity in coherent optical communication systems with ultralow complexity. To validate the proposed methodology, we carry out an experiment of ten-channel wavelength division multiplexing (WDM) transmission with 64 Gbaud polarization-division-multiplexed 16-ary quadrature amplitude modulation (16-QAM) signals. Co-LSTM and DBP achieve a comparable performance of nonlinear mitigation. However, the complexity of Co-LSTM with a simplified mode is almost independent of the transmission distance, which is much lower than that of the DBP. The proposed Co-LSTM methodology presents an attractive approach for low complexity nonlinearity mitigation with neural networks.

preprint2021arXiv

VIVO: Visual Vocabulary Pre-Training for Novel Object Captioning

It is highly desirable yet challenging to generate image captions that can describe novel objects which are unseen in caption-labeled training data, a capability that is evaluated in the novel object captioning challenge (nocaps). In this challenge, no additional image-caption training data, other thanCOCO Captions, is allowed for model training. Thus, conventional Vision-Language Pre-training (VLP) methods cannot be applied. This paper presents VIsual VOcabulary pretraining (VIVO) that performs pre-training in the absence of caption annotations. By breaking the dependency of paired image-caption training data in VLP, VIVO can leverage large amounts of paired image-tag data to learn a visual vocabulary. This is done by pre-training a multi-layer Transformer model that learns to align image-level tags with their corresponding image region features. To address the unordered nature of image tags, VIVO uses a Hungarian matching loss with masked tag prediction to conduct pre-training. We validate the effectiveness of VIVO by fine-tuning the pre-trained model for image captioning. In addition, we perform an analysis of the visual-text alignment inferred by our model. The results show that our model can not only generate fluent image captions that describe novel objects, but also identify the locations of these objects. Our single model has achieved new state-of-the-art results on nocaps and surpassed the human CIDEr score.

preprint2021arXiv

Weak phases and CP-symmetry tests in sequential decays of entangled double-strange baryons

Using a sample of $1.31\times10^9$ $J/ψ$ events collected with the BESIII detector at the electron-positron collider BEPCII, we analyse the full $J/ψ\to$ $Ξ^-\overlineΞ^+$, $Ξ^-\to Λπ^-$, $Λ\to pπ^-$, $\overlineΞ^+\to\overlineΛπ^+$, $\overlineΛ\to\overline{p}π^+$ decay chain. A new method, exploiting the fact that the $Ξ^-\overlineΞ^+$ pair is entangled and sequentially decaying, and where the complete decay chains are reconstructed, is applied for the first time. This enables precision measurements of the decay parameters for the $Ξ^-\toΛπ^-$ decay ($α_Ξ$, $ϕ_Ξ$) as well as the $\overlineΞ^+\to\overlineΛπ^+$ decay ($\overlineα_Ξ$, $\overlineϕ_Ξ$). From the decay parameters, two independent CP tests were performed, quantified by the observables $A_{\rm CP}^Ξ$ and $Δϕ_Ξ$. Our results, $A_{\rm CP}^Ξ$ = $(6.0\pm13.4\pm5.6)\times10^{-3}$ and $Δϕ_Ξ= (-4.8\pm13.7\pm2.9)\times10^{-3}~{\rm rad}$, are consistent with CP symmetry. Furthermore, our method enables a separation of strong and weak $Ξ\toΛπ$ decay amplitudes. This results in the first direct measurement of the weak phase difference for any baryon decay. The result is found to be $(ξ_{P} - ξ_{S}) = (1.2\pm3.4\pm0.8)\times10^{-2}$ rad and is one of the most precise tests of CP symmetry for strange baryons. The strong phase difference is measured to be $(δ_P - δ_S) = (-4.0\pm3.3\pm1.7)\times10^{-2}$ rad. In addition, we provide an independent measurement of the recently debated $Λ$ decay parameter, $α_Λ = 0.757 \pm 0.011 \pm 0.008 $. The $Λ\overlineΛ$ asymmetry is measured to be $A_{\rm CP}^Λ = (-3.7\pm11.7\pm9.0)\times10^{-3}$.

preprint2020arXiv

A numerical approach for hybrid reliability analysis of structures under mixed uncertainties using the uncertainty theory

This paper presents a novel numerical method for the hybrid reliability analysis by using the uncertainty theory. Aleatory uncertainty and epistemic uncertainty are considered simultaneously in this method. Epistemic uncertainty is characterized by the uncertainty theory, and the effect of epistemic uncertainty is quantified by the sub-additive uncertain measure. Then, under the framework of the chance theory which can be interpreted as the combination of the probability theory and the uncertainty theory, a general uncertainty quantification model is established to deal with the hybrid reliability analysis problem, then the corresponding reliability metric is defined. After that, to improve the feasibility of the proposed model, by utilizing the polar coordinate transformation based dimension reduction method, a numerical analysis method for the hybrid reliability model are provided. At last, several application cases are presented to prove the effectiveness of the proposed method for the reliability analysis under hybrid uncertainty. The comparisons between the results of the proposed method and the Monte Carlo simulation also illustrate the merit of this method.

preprint2020arXiv

A Posteriori Error Estimates for Adaptive QM/MM Coupling Methods

Hybrid quantum/molecular mechanics models (QM/MM methods) are widely used in material and molecular simulations when MM models do not provide sufficient accuracy but pure QM models are computationally prohibitive. Adaptive QM/MM coupling methods feature on-the-fly classification of atoms during the simulation, allowing the QM and MM subsystems to be updated as needed. In this work, we propose such an adaptive QM/MM method for material defect simulations based on a new residual based it a posteriori error estimator, which provides both lower and upper bounds for the true error. We validate the analysis and illustrate the effectiveness of the new scheme on numerical simulations for material defects.

preprint2020arXiv

A Reduced Study for Nematic Equilibria on Two-Dimensional Polygons

We study reduced nematic equilibria on regular two-dimensional polygons with Dirichlet tangent boundary conditions, in a reduced two-dimensional Landau-de Gennes framework, discussing their relevance in the full three-dimensional framework too. We work at a fixed temperature and study the reduced stable equilibria in terms of the edge length, $λ$ of the regular polygon, $E_K$ with $K$ edges. We analytically compute a novel "ring solution" in the $λ\to 0$ limit, with a unique point defect at the centre of the polygon for $K \neq 4$. The ring solution is unique. For sufficiently large $λ$, we deduce the existence of at least $\left[K/2 \right]$ classes of stable equilibria and numerically compute bifurcation diagrams for reduced equilibria on a pentagon and hexagon, as a function of $λ^2$, thus illustrating the effects of geometry on the structure, locations and dimensionality of defects in this framework.

preprint2020arXiv

A Single Stream Network for Robust and Real-time RGB-D Salient Object Detection

Existing RGB-D salient object detection (SOD) approaches concentrate on the cross-modal fusion between the RGB stream and the depth stream. They do not deeply explore the effect of the depth map itself. In this work, we design a single stream network to directly use the depth map to guide early fusion and middle fusion between RGB and depth, which saves the feature encoder of the depth stream and achieves a lightweight and real-time model. We tactfully utilize depth information from two perspectives: (1) Overcoming the incompatibility problem caused by the great difference between modalities, we build a single stream encoder to achieve the early fusion, which can take full advantage of ImageNet pre-trained backbone model to extract rich and discriminative features. (2) We design a novel depth-enhanced dual attention module (DEDA) to efficiently provide the fore-/back-ground branches with the spatially filtered features, which enables the decoder to optimally perform the middle fusion. Besides, we put forward a pyramidally attended feature extraction module (PAFE) to accurately localize the objects of different scales. Extensive experiments demonstrate that the proposed model performs favorably against most state-of-the-art methods under different evaluation metrics. Furthermore, this model is 55.5\% lighter than the current lightest model and runs at a real-time speed of 32 FPS when processing a $384 \times 384$ image.

preprint2020arXiv

Abnormal activity capture from passenger flow of elevator based on unsupervised learning and fine-grained multi-label recognition

We present a work-flow which aims at capturing residents' abnormal activities through the passenger flow of elevator in multi-storey residence buildings. Camera and sensors (hall sensor, photoelectric sensor, gyro, accelerometer, barometer, and thermometer) with internet connection are mounted in elevator to collect image and data. Computer vision algorithms such as instance segmentation, multi-label recognition, embedding and clustering are applied to generalize passenger flow of elevator, i.e. how many people and what kinds of people get in and out of the elevator on each floor. More specifically in our implementation we propose GraftNet, a solution for fine-grained multi-label recognition task, to recognize human attributes, e.g. gender, age, appearance, and occupation. Then anomaly detection of unsupervised learning is hierarchically applied on the passenger flow data to capture abnormal or even illegal activities of the residents which probably bring safety hazard, e.g. drug dealing, pyramid sale gathering, prostitution, and over crowded residence. Experiment shows effects are there, and the captured records will be directly reported to our customer(property managers) for further confirmation.

preprint2020arXiv

AIM 2020 Challenge on Efficient Super-Resolution: Methods and Results

This paper reviews the AIM 2020 challenge on efficient single image super-resolution with focus on the proposed solutions and results. The challenge task was to super-resolve an input image with a magnification factor x4 based on a set of prior examples of low and corresponding high resolution images. The goal is to devise a network that reduces one or several aspects such as runtime, parameter count, FLOPs, activations, and memory consumption while at least maintaining PSNR of MSRResNet. The track had 150 registered participants, and 25 teams submitted the final results. They gauge the state-of-the-art in efficient single image super-resolution.

preprint2020arXiv

Aligning Partially Overlapping Point Sets: an Inner Approximation Algorithm

Aligning partially overlapping point sets where there is no prior information about the value of the transformation is a challenging problem in computer vision. To achieve this goal, we first reduce the objective of the robust point matching algorithm to a function of a low dimensional variable. The resulting function, however, is only concave over a finite region including the feasible region. To cope with this issue, we employ the inner approximation optimization algorithm which only operates within the region where the objective function is concave. Our algorithm does not need regularization on transformation, and thus can handle the situation where there is no prior information about the values of the transformations. Our method is also $ε-$globally optimal and thus is guaranteed to be robust. Moreover, its most computationally expensive subroutine is a linear assignment problem which can be efficiently solved. Experimental results demonstrate the better robustness of the proposed method over state-of-the-art algorithms. Our method is also efficient when the number of transformation parameters is small.

preprint2020arXiv

An Interstate Trips Analysis during COVID-19 in the United States

The worldwide outbreak of COVID-19 has posed a dire threat to the public. Human mobility has changed in various ways over the course of the pandemic. Despite current studies on common mobility metrics, research specifically on state-to-state mobility is very limited. By leveraging the mobile phone location data from over 100 million anonymous devices, we estimate the population flow between all states in the United States. We first analyze the temporal pattern and spatial differences of between-state flow from January 1, 2020 to May 15, 2020. Then, with repeated measures ANOVA and post-hoc analysis, we discern different time-course patterns of between-state population flow by pandemic severity groups. A further analysis shows moderate to high correlation between the flow reduction and the pandemic severity, the strength of which varies with different policies. This paper is promising in predicting imported cases.

preprint2020arXiv

Asymptotic behavior of the basic reproduction ratio for periodic reaction-diffusion systems

This paper is devoted to the study of asymptotic behavior of the basic reproduction ratio for periodic reaction-diffusion systems in the case of small and large diffusion coefficients. We first establish the continuity of the basic reproduction ratio with respect to parameters by developing the theory of resolvent positive operators. Then we investigate the limiting profile of the principal eigenvalue of an associated periodic eigenvalue problem for large diffusion coefficients. We then obtain the asymptotic behavior of the basic reproduction ratio as the diffusion coefficients go to zero and infinity, respectively. We also investigate the limiting behavior of positive periodic solution for periodic and cooperative reaction-diffusion systems with the Neumann boundary condition when the diffusion coefficients are large enough. Finally, we apply these results to a reaction-diffusion model of Zika virus transmission.

preprint2020arXiv

Blended Ghost Force Correction Method for 3D Crystalline Defects

Atomistic/continuum coupling method is a class of multiscale computational method for the efficient simulation of crystalline defects. The recently developed blended ghost force correction (BGFC) method combines the efficiency of blending methods and the accuracy of QNL type methods. BGFC method can be applied to multi-body interaction potentials and general interfaces. In this paper, we present the formulation, implementation and analysis of the BGFC method in three dimensions. In particular, we focus on the difference and connection with other blending variants, such as energy based blended quasi-continuum method (BQCE) and force based blended quasi-continuum method (BQCF). The theoretical results are justified by a few benchmark numerical experiments with point defects and microcrack in the three dimensional FCC lattice.

preprint2020arXiv

Blind Face Restoration via Deep Multi-scale Component Dictionaries

Recent reference-based face restoration methods have received considerable attention due to their great capability in recovering high-frequency details on real low-quality images. However, most of these methods require a high-quality reference image of the same identity, making them only applicable in limited scenes. To address this issue, this paper suggests a deep face dictionary network (termed as DFDNet) to guide the restoration process of degraded observations. To begin with, we use K-means to generate deep dictionaries for perceptually significant face components (\ie, left/right eyes, nose and mouth) from high-quality images. Next, with the degraded input, we match and select the most similar component features from their corresponding dictionaries and transfer the high-quality details to the input via the proposed dictionary feature transfer (DFT) block. In particular, component AdaIN is leveraged to eliminate the style diversity between the input and dictionary features (\eg, illumination), and a confidence score is proposed to adaptively fuse the dictionary feature to the input. Finally, multi-scale dictionaries are adopted in a progressive manner to enable the coarse-to-fine restoration. Experiments show that our proposed method can achieve plausible performance in both quantitative and qualitative evaluation, and more importantly, can generate realistic and promising results on real degraded images without requiring an identity-belonging reference. The source code and models are available at \url{https://github.com/csxmli2016/DFDNet}.

preprint2020arXiv

Boosting Weakly Supervised Object Detection with Progressive Knowledge Transfer

In this paper, we propose an effective knowledge transfer framework to boost the weakly supervised object detection accuracy with the help of an external fully-annotated source dataset, whose categories may not overlap with the target domain. This setting is of great practical value due to the existence of many off-the-shelf detection datasets. To more effectively utilize the source dataset, we propose to iteratively transfer the knowledge from the source domain by a one-class universal detector and learn the target-domain detector. The box-level pseudo ground truths mined by the target-domain detector in each iteration effectively improve the one-class universal detector. Therefore, the knowledge in the source dataset is more thoroughly exploited and leveraged. Extensive experiments are conducted with Pascal VOC 2007 as the target weakly-annotated dataset and COCO/ImageNet as the source fully-annotated dataset. With the proposed solution, we achieved an mAP of $59.7\%$ detection performance on the VOC test set and an mAP of $60.2\%$ after retraining a fully supervised Faster RCNN with the mined pseudo ground truths. This is significantly better than any previously known results in related literature and sets a new state-of-the-art of weakly supervised object detection under the knowledge transfer setting. Code: \url{https://github.com/mikuhatsune/wsod_transfer}.

preprint2020arXiv

Branch-Cooperative OSNet for Person Re-Identification

Multi-branch is extensively studied for learning rich feature representation for person re-identification (Re-ID). In this paper, we propose a branch-cooperative architecture over OSNet, termed BC-OSNet, for person Re-ID. By stacking four cooperative branches, namely, a global branch, a local branch, a relational branch and a contrastive branch, we obtain powerful feature representation for person Re-ID. Extensive experiments show that the proposed BC-OSNet achieves state-of-art performance on the three popular datasets, including Market-1501, DukeMTMC-reID and CUHK03. In particular, it achieves mAP of 84.0% and rank-1 accuracy of 87.1% on the CUHK03_labeled.

preprint2020arXiv

Construction of a minimum energy path for the VT flash model by an exponential time differencing scheme with the string method

Phase equilibrium calculation, also known as flash calculation, plays significant roles in various aspects of petroleum and chemical industries. Since Michelsen proposed his milestone studies in 1982, through several decades of development, the current research interest on flash calculation has been shifted from accuracy to efficiency, but the ultimate goal remains the same focusing on estimation of the equilibrium phase amounts and phase compositions under the given variable specification. However, finding the transition route and its related saddle points are very often helpful to study the evolution of phase change and partition. Motivated by this, in this study we apply the string method to find the minimum energy paths and saddle points information of a single-component VT flash model with the Peng-Robinson equation of state. As the system has strong stiffness, common ordinary differential equation solvers have their limitations. To overcome these issues, a Rosenbrock-type exponential time differencing scheme is employed to reduce the computational difficulty caused by the high stiffness of the investigated system. In comparison with the published results and experimental data, the proposed numerical algorithm not only shows good feasibility and accuracy on phase equilibrium calculation, but also successfully calculates the minimum energy path and and saddle point of the single-component VT flash model with strong stiffness.

preprint2020arXiv

Deep CNNs Meet Global Covariance Pooling: Better Representation and Generalization

Compared with global average pooling in existing deep convolutional neural networks (CNNs), global covariance pooling can capture richer statistics of deep features, having potential for improving representation and generalization abilities of deep CNNs. However, integration of global covariance pooling into deep CNNs brings two challenges: (1) robust covariance estimation given deep features of high dimension and small sample size; (2) appropriate usage of geometry of covariances. To address these challenges, we propose a global Matrix Power Normalized COVariance (MPN-COV) Pooling. Our MPN-COV conforms to a robust covariance estimator, very suitable for scenario of high dimension and small sample size. It can also be regarded as Power-Euclidean metric between covariances, effectively exploiting their geometry. Furthermore, a global Gaussian embedding network is proposed to incorporate first-order statistics into MPN-COV. For fast training of MPN-COV networks, we implement an iterative matrix square root normalization, avoiding GPU unfriendly eigen-decomposition inherent in MPN-COV. Additionally, progressive 1x1 convolutions and group convolution are introduced to compress covariance representations. The proposed methods are highly modular, readily plugged into existing deep CNNs. Extensive experiments are conducted on large-scale object classification, scene categorization, fine-grained visual recognition and texture classification, showing our methods outperform the counterparts and obtain state-of-the-art performance.

preprint2020arXiv

Domain Adaptive Object Detection via Asymmetric Tri-way Faster-RCNN

Conventional object detection models inevitably encounter a performance drop as the domain disparity exists. Unsupervised domain adaptive object detection is proposed recently to reduce the disparity between domains, where the source domain is label-rich while the target domain is label-agnostic. The existing models follow a parameter shared siamese structure for adversarial domain alignment, which, however, easily leads to the collapse and out-of-control risk of the source domain and brings negative impact to feature adaption. The main reason is that the labeling unfairness (asymmetry) between source and target makes the parameter sharing mechanism unable to adapt. Therefore, in order to avoid the source domain collapse risk caused by parameter sharing, we propose an asymmetric tri-way Faster-RCNN (ATF) for domain adaptive object detection. Our ATF model has two distinct merits: 1) A ancillary net supervised by source label is deployed to learn ancillary target features and simultaneously preserve the discrimination of source domain, which enhances the structural discrimination (object classification vs. bounding box regression) of domain alignment. 2) The asymmetric structure consisting of a chief net and an independent ancillary net essentially overcomes the parameter sharing aroused source risk collapse. The adaption safety of the proposed ATF detector is guaranteed. Extensive experiments on a number of datasets, including Cityscapes, Foggy-cityscapes, KITTI, Sim10k, Pascal VOC, Clipart and Watercolor, demonstrate the SOTA performance of our method.

preprint2020arXiv

Domain Private and Agnostic Feature for Modality Adaptive Face Recognition

Heterogeneous face recognition is a challenging task due to the large modality discrepancy and insufficient cross-modal samples. Most existing works focus on discriminative feature transformation, metric learning and cross-modal face synthesis. However, the fact that cross-modal faces are always coupled by domain (modality) and identity information has received little attention. Therefore, how to learn and utilize the domain-private feature and domain-agnostic feature for modality adaptive face recognition is the focus of this work. Specifically, this paper proposes a Feature Aggregation Network (FAN), which includes disentangled representation module (DRM), feature fusion module (FFM) and adaptive penalty metric (APM) learning session. First, in DRM, two subnetworks, i.e. domain-private network and domain-agnostic network are specially designed for learning modality features and identity features, respectively. Second, in FFM, the identity features are fused with domain features to achieve cross-modal bi-directional identity feature transformation, which, to a large extent, further disentangles the modality information and identity information. Third, considering that the distribution imbalance between easy and hard pairs exists in cross-modal datasets, which increases the risk of model bias, the identity preserving guided metric learning with adaptive hard pairs penalization is proposed in our FAN. The proposed APM also guarantees the cross-modality intra-class compactness and inter-class separation. Extensive experiments on benchmark cross-modal face datasets show that our FAN outperforms SOTA methods.

preprint2020arXiv

Dual Adversarial Network: Toward Real-world Noise Removal and Noise Generation

Real-world image noise removal is a long-standing yet very challenging task in computer vision. The success of deep neural network in denoising stimulates the research of noise generation, aiming at synthesizing more clean-noisy image pairs to facilitate the training of deep denoisers. In this work, we propose a novel unified framework to simultaneously deal with the noise removal and noise generation tasks. Instead of only inferring the posteriori distribution of the latent clean image conditioned on the observed noisy image in traditional MAP framework, our proposed method learns the joint distribution of the clean-noisy image pairs. Specifically, we approximate the joint distribution with two different factorized forms, which can be formulated as a denoiser mapping the noisy image to the clean one and a generator mapping the clean image to the noisy one. The learned joint distribution implicitly contains all the information between the noisy and clean images, avoiding the necessity of manually designing the image priors and noise assumptions as traditional. Besides, the performance of our denoiser can be further improved by augmenting the original training dataset with the learned generator. Moreover, we propose two metrics to assess the quality of the generated noisy image, for which, to the best of our knowledge, such metrics are firstly proposed along this research line. Extensive experiments have been conducted to demonstrate the superiority of our method over the state-of-the-arts both in the real noise removal and generation tasks. The training and testing code is available at https://github.com/zsyOAOA/DANet.

preprint2020arXiv

Erratum to "Measurement of the $e^+e^-\toπ^+π^-$ cross section between 600 and 900 MeV using initial state radiation"

In Phys. Lett. B 753, 629-638 (2016) [arXiv:1507.08188] the BESIII collaboration published a cross section measurement of the process $e^+e^-\to π^+ π^-$ in the energy range between 600 and 900 MeV. In this erratum we report a corrected evaluation of the statistical errors in terms of a fully propagated covariance matrix. The correction also yields a reduced statistical uncertainty for the hadronic vacuum polarization contribution to the anomalous magnetic moment of the muon, which now reads as $a_μ^{ππ\mathrm{, LO}}(600 - 900\,\mathrm{MeV}) = (368.2 \pm 1.5_{\rm stat} \pm 3.3_{\rm syst})\times 10^{-10}$. The central values of the cross section measurement and of $a_μ^{ππ\mathrm{, LO}}$, as well as the systematic uncertainties remain unchanged.

preprint2020arXiv

Hard Negative Samples Emphasis Tracker without Anchors

Trackers based on Siamese network have shown tremendous success, because of their balance between accuracy and speed. Nevertheless, with tracking scenarios becoming more and more sophisticated, most existing Siamese-based approaches ignore the addressing of the problem that distinguishes the tracking target from hard negative samples in the tracking phase. The features learned by these networks lack of discrimination, which significantly weakens the robustness of Siamese-based trackers and leads to suboptimal performance. To address this issue, we propose a simple yet efficient hard negative samples emphasis method, which constrains Siamese network to learn features that are aware of hard negative samples and enhance the discrimination of embedding features. Through a distance constraint, we force to shorten the distance between exemplar vector and positive vectors, meanwhile, enlarge the distance between exemplar vector and hard negative vectors. Furthermore, we explore a novel anchor-free tracking framework in a per-pixel prediction fashion, which can significantly reduce the number of hyper-parameters and simplify the tracking process by taking full advantage of the representation of convolutional neural network. Extensive experiments on six standard benchmark datasets demonstrate that the proposed method can perform favorable results against state-of-the-art approaches.

preprint2020arXiv

Hierarchical Bi-Directional Feature Perception Network for Person Re-Identification

Previous Person Re-Identification (Re-ID) models aim to focus on the most discriminative region of an image, while its performance may be compromised when that region is missing caused by camera viewpoint changes or occlusion. To solve this issue, we propose a novel model named Hierarchical Bi-directional Feature Perception Network (HBFP-Net) to correlate multi-level information and reinforce each other. First, the correlation maps of cross-level feature-pairs are modeled via low-rank bilinear pooling. Then, based on the correlation maps, Bi-directional Feature Perception (BFP) module is employed to enrich the attention regions of high-level feature, and to learn abstract and specific information in low-level feature. And then, we propose a novel end-to-end hierarchical network which integrates multi-level augmented features and inputs the augmented low- and middle-level features to following layers to retrain a new powerful network. What's more, we propose a novel trainable generalized pooling, which can dynamically select any value of all locations in feature maps to be activated. Extensive experiments implemented on the mainstream evaluation datasets including Market-1501, CUHK03 and DukeMTMC-ReID show that our method outperforms the recent SOTA Re-ID models.

preprint2020arXiv

How different age groups responded to the COVID-19 pandemic in terms of mobility behaviors: a case study of the United States

The rapid spread of COVID-19 has affected thousands of people from different socio-demographic groups all over the country. A decisive step in preventing or slowing the outbreak is the use of mobility interventions, such as government stay-at-home orders. However, different socio-demographic groups might have different responses to these orders and regulations. In this paper, we attempt to fill the current gap in the literature by examining how different communities with different age groups performed social distancing by following orders such as the national emergency declaration on March 13, as well as how fast they started changing their behavior after the regulations were imposed. For this purpose, we calculated the behavior changes of people in different mobility metrics, such as percentage of people staying home during the study period (March, April, and May 2020), in different age groups in comparison to the days before the pandemic (January and February 2020), by utilizing anonymized and privacy-protected mobile device data. Our study indicates that senior communities outperformed younger communities in terms of their behavior change. Senior communities not only had a faster response to the outbreak in comparison to young communities, they also had better performance consistency during the pandemic.

preprint2020arXiv

Human Mobility Trends during the COVID-19 Pandemic in the United States

In March of this year, COVID-19 was declared a pandemic and it continues to threaten public health. This global health crisis imposes limitations on daily movements, which have deteriorated every sector in our society. Understanding public reactions to the virus and the non-pharmaceutical interventions should be of great help to fight COVID-19 in a strategic way. We aim to provide tangible evidence of the human mobility trends by comparing the day-by-day variations across the U.S. Large-scale public mobility at an aggregated level is observed by leveraging mobile device location data and the measures related to social distancing. Our study captures spatial and temporal heterogeneity as well as the sociodemographic variations regarding the pandemic propagation and the non-pharmaceutical interventions. All mobility metrics adapted capture decreased public movements after the national emergency declaration. The population staying home has increased in all states and becomes more stable after the stay-at-home order with a smaller range of fluctuation. There exists overall mobility heterogeneity between the income or population density groups. The public had been taking active responses, voluntarily staying home more, to the in-state confirmed cases while the stay-at-home orders stabilize the variations. The study suggests that the public mobility trends conform with the government message urging to stay home. We anticipate our data-driven analysis offers integrated perspectives and serves as evidence to raise public awareness and, consequently, reinforce the importance of social distancing while assisting policymakers.

preprint2020arXiv

Inclusive charged and neutral particle multiplicity distributions in $χ_{cJ}$ and $J/ψ$ decays

Using a sample of 106 million $ψ(3686)$ decays, $ψ(3686) \to γχ_{cJ} (J = 0, 1, 2)$ and $ψ(3686) \to γχ_{cJ}, χ_{cJ} \to γJ/ψ$ $(J = 1, 2)$ events are utilized to study inclusive $χ_{cJ} \to$ anything, $χ_{cJ} \to$ hadrons, and $J/ψ\to$ anything distributions, including distributions of the number of charged tracks, electromagnetic calorimeter showers, and $π^0$s, and to compare them with distributions obtained from the BESIII Monte Carlo simulation. Information from each Monte Carlo simulated decay event is used to construct matrices connecting the detected distributions to the input predetection "produced" distributions. Assuming these matrices also apply to data, they are used to predict the analogous produced distributions of the decay events. Using these, the charged particle multiplicities are compared with results from MARK I. Further, comparison of the distributions of the number of photons in data with those in Monte Carlo simulation indicates that G-parity conservation should be taken into consideration in the simulation.

preprint2020arXiv

Isolated singularities of solutions to the Yamabe equation in dimension $6$

We study the asymptotic behavior of local solutions to the Yamabe equation near an isolated singularity, when the metric is not conformally flat. We prove that, in dimension $6$, any solution is asymptotically close to a Fowler solution, which is an extension of the same result for lower dimensions by F.C. Marques in 2008.

preprint2020arXiv

Label Propagation with Augmented Anchors: A Simple Semi-Supervised Learning baseline for Unsupervised Domain Adaptation

Motivated by the problem relatedness between unsupervised domain adaptation (UDA) and semi-supervised learning (SSL), many state-of-the-art UDA methods adopt SSL principles (e.g., the cluster assumption) as their learning ingredients. However, they tend to overlook the very domain-shift nature of UDA. In this work, we take a step further to study the proper extensions of SSL techniques for UDA. Taking the algorithm of label propagation (LP) as an example, we analyze the challenges of adopting LP to UDA and theoretically analyze the conditions of affinity graph/matrix construction in order to achieve better propagation of true labels to unlabeled instances. Our analysis suggests a new algorithm of Label Propagation with Augmented Anchors (A$^2$LP), which could potentially improve LP via generation of unlabeled virtual instances (i.e., the augmented anchors) with high-confidence label predictions. To make the proposed A$^2$LP useful for UDA, we propose empirical schemes to generate such virtual instances. The proposed schemes also tackle the domain-shift challenge of UDA by alternating between pseudo labeling via A$^2$LP and domain-invariant feature learning. Experiments show that such a simple SSL extension improves over representative UDA methods of domain-invariant feature learning, and could empower two state-of-the-art methods on benchmark UDA datasets. Our results show the value of further investigation on SSL techniques for UDA problems.

preprint2020arXiv

Largely enhanced photogalvanic effects in the phosphorene photodetector by strain-increased device asymmetry

Photogalvanic effect (PGE) occurring in noncentrosymmetric materials enables the generation of the open-circuit voltage that is much larger than the bandgap, making it rather attractive in solar cells. However, the magnitude of the PGE photocurrent is usually small, which severely hampers its practical application. Here we propose a mechanism to largely enhance the PGE photocurrent by mechanical strain based on the quantum transport simulations for the two-dimensional nickel-phosphorene-nickel photodetector. Broadband PGE photocurrent governed by the Cs noncentrosymmetry is generated at zero bias under the illumination of linearly polarized light. The photocurrent depends linearly on the device asymmetry, while nonlinearly on the optical absorption. By applying the appropriate mechanical tension stress on the phosphorene, the photocurrent can be substantially enhanced by up to 3 orders of magnitude, which is primarily ascribed to the largely increased device asymmetry. The change in the optical absorption in some cases can also play a critical role in tuning the photocurrent due to the nonlinear dependence. Moreover, the photocurrent can even be further enhanced by the mechanical bending, mainly owing to the considerably enhanced device asymmetry. Our results reveal the dependence of the PGE photocurrent on the device asymmetry and absorption in transport process through a device, and also explore the potentials of the PGE in the self-powered low-dimensional flexible optoelectronics.

preprint2020arXiv

Measurement of Singly Cabibbo-Suppressed Decays $D \to ωππ$

Using 2.93 fb$^{-1}$ of $e^{+}e^{-}$ collision data taken at a center-of-mass energy of 3.773 GeV by the BESIII detector at the BEPCII, we measure the branching fractions of the singly Cabibbo-suppressed decays $D \to ωππ$ to be $\mathcal{B}(D^0 \to ωπ^+π^-) = (1.33 \pm 0.16 \pm 0.12)\times 10^{-3}$ and $\mathcal{B}(D^+ \to ωπ^+π^0) =(3.87 \pm 0.83 \pm 0.25)\times 10^{-3}$, where the first uncertainties are statistical and the second ones systematic. The statistical significances are $12.9σ$ and $7.7 σ$, respectively. The precision of $\mathcal{B}(D^0 \to ωπ^+π^-)$ is improved by a factor of 2.1 over the CLEO measurement, and $\mathcal{B}(D^+ \to ωπ^+π^0)$ is measured for the first time. No significant signal of $\mathcal{B}(D^0 \to ωπ^0π^0)$ is observed, and the upper limit on the branching fraction is $\mathcal{B}(D^0 \to ωπ^0π^0) < 1.10 \times 10^{-3}$ at the $90\%$ confidence level. The branching fractions of $D\to ηππ$ are also measured and consistent with existing results.

preprint2020arXiv

Melanoma Diagnosis with Spatio-Temporal Feature Learning on Sequential Dermoscopic Images

Existing studies for automated melanoma diagnosis are based on single-time point images of lesions. However, melanocytic lesions de facto are progressively evolving and, moreover, benign lesions can progress into malignant melanoma. Ignoring cross-time morphological changes of lesions thus may lead to misdiagnosis in borderline cases. Based on the fact that dermatologists diagnose ambiguous skin lesions by evaluating the dermoscopic changes over time via follow-up examination, in this study, we propose an automated framework for melanoma diagnosis using sequential dermoscopic images. To capture the spatio-temporal characterization of dermoscopic evolution, we construct our model in a two-stream network architecture which capable of simultaneously learning appearance representations of individual lesions while performing temporal reasoning on both raw pixels difference and abstract features difference. We collect 184 cases of serial dermoscopic image data, which consists of histologically confirmed 92 benign lesions and 92 melanoma lesions, to evaluate the effectiveness of the proposed method. Our model achieved AUC of 74.34%, which is ~8% higher than that of only using single images and ~6% higher than the widely used sequence learning model based on LSTM.

preprint2020arXiv

Modeling indoor-level non-pharmaceutical interventions during the COVID-19 pandemic: a pedestrian dynamics-based microscopic simulation approach

Mathematical modeling of epidemic spreading has been widely adopted to estimate the threats of epidemic diseases (i.e., the COVID-19 pandemic) as well as to evaluate epidemic control interventions. The indoor place is considered to be a significant epidemic spreading risk origin, but existing widely-used epidemic spreading models are usually limited for indoor places since the dynamic physical distance changes between people are ignored, and the empirical features of the essential and non-essential travel are not differentiated. In this paper, we introduce a pedestrian-based epidemic spreading model that is capable of modeling indoor transmission risks of diseases during people's social activities. Taking advantage of the before-and-after mobility data from the University of Maryland COVID-19 Impact Analysis Platform, it's found that people tend to spend more time in grocery stores once their travel frequencies are restricted to a low level. In other words, an increase in dwell time could balance the decrease in travel frequencies and satisfy people's demand. Based on the pedestrian-based model and the empirical evidence, combined non-pharmaceutical interventions from different operational levels are evaluated. Numerical simulations show that restrictions on people's travel frequency and open-hours of indoor places may not be universally effective in reducing average infection risks for each pedestrian who visit the place. Entry limitations can be a widely effective alternative, whereas the decision-maker needs to balance the decrease in risky contacts and the increase in queue length outside the place that may impede people from fulfilling their travel needs.

preprint2020arXiv

Nonuniversal Entanglement Level Statistics in Projection-driven Quantum Circuits

We study the level-spacing statistics in the entanglement spectrum of output states of random universal quantum circuits where qubits are subject to a finite probability of projection to the computational basis at each time step. We encounter two phase transitions with increasing projection rate: The first is the volume-to-area law transition observed in quantum circuits with projective measurements; The second separates the pure Poisson level statistics phase at large projective measurement rates from a regime of residual level repulsion in the entanglement spectrum within the area-law phase, characterized by non-universal level spacing statistics that interpolates between the Wigner-Dyson and Poisson distributions. By applying a tensor network contraction algorithm introduced in Ref. [1] to the circuit spacetime, we identify this second projective-measurement-driven transition as a percolation transition of entangled bonds. The same behavior is observed in both circuits of random two-qubit unitaries and circuits of universal gate sets, including the set implemented by Google in its Sycamore circuits.

preprint2020arXiv

Observation of a resonant structure in $e^{+}e^{-} \to ωη$ and another in $e^{+}e^{-} \to ωπ^{0}$ at center-of-mass energies between 2.00 and 3.08 GeV

Born cross sections for the processes $e^+e^- \to ωη$ and $e^+e^- \to ωπ^{0}$ have been determined for center-of-mass energies between 2.00 and 3.08 GeV with the BESIII detector at the BEPCII collider. The results obtained in this work are consistent with previous measurements but with improved precision. Two resonant structures are observed. In the $e^{+}e^{-} \to ωη$ cross sections, a resonance with a mass of $(2179 \pm 21 \pm 3)\text{MeV}/c^2$ and a width of $(89 \pm 28 \pm 5)\text{MeV}$ is observed with a significance of 6.1$σ$. Its properties are consistent with the $ϕ(2170)$. In the $e^{+}e^{-} \toωπ^{0}$ cross sections, a resonance denoted $Y(2040)$ is observed with a significance of more than 10$σ$. Its mass and width are determined to be $(2034 \pm 13 \pm 9)\text{MeV}/c^2$ and $(234 \pm 30 \pm 25)\text{MeV}$, respectively, where the first uncertainties are statistical and the second ones are systematic.

preprint2020arXiv

Observation of a structure in $e^+e^- \to ϕη^{\prime}$ at $\sqrt{s}$ from 2.05 to 3.08 GeV

The process $e^{+}e^{-} \to ϕη^{\prime}$ has been studied for the first time in detail using data sample collected with the BESIII detector at the BEPCII collider at center of mass energies from 2.05 to 3.08 GeV. A resonance with quantum numbers $J^{PC}=1^{--}$ is observed with mass $M$ = (2177.5 $\pm$ 4.8 (stat) $\pm$ 19.5 (syst)) MeV/${ \it{c}^{\mathrm{2}}}$ and width $Γ$ = (149.0 $\pm$ 15.6 (stat) $\pm$ 8.9 (syst)) MeV with a statistical significance larger than 10$σ$. The observed structure could be identified with the $ϕ(2170)$, then the ratio of partial width between the $ϕη^{\prime}$ by BESIII and $ϕη$ by BABAR is ($\mathcal{B}^{R}_{ϕη}Γ^{R}_{ee})/{(\mathcal{B}^{R}_{ϕη^{\prime}}Γ^{R}_{ee})}$ = 0.23 $\pm$ 0.10 (stat) $\pm$ 0.18 (syst), which is smaller than the prediction of the $s\bar{s}g$ hybrid models by several orders of magnitude.

preprint2020arXiv

Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks

Large-scale pre-training methods of learning cross-modal representations on image-text pairs are becoming popular for vision-language tasks. While existing methods simply concatenate image region features and text features as input to the model to be pre-trained and use self-attention to learn image-text semantic alignments in a brute force manner, in this paper, we propose a new learning method Oscar (Object-Semantics Aligned Pre-training), which uses object tags detected in images as anchor points to significantly ease the learning of alignments. Our method is motivated by the observation that the salient objects in an image can be accurately detected, and are often mentioned in the paired text. We pre-train an Oscar model on the public corpus of 6.5 million text-image pairs, and fine-tune it on downstream tasks, creating new state-of-the-arts on six well-established vision-language understanding and generation tasks.

preprint2020arXiv

Quantifying the influence of inter-county mobility patterns on the COVID-19 outbreak in the United States

As a highly infectious respiratory disease, COVID-19 has become a pandemic that threatens global health. Without an effective treatment, non-pharmaceutical interventions, such as travel restrictions, have been widely promoted to mitigate the outbreak. Current studies analyze mobility metrics such as travel distance; however, there is a lack of research on interzonal travel flow and its impact on the pandemic. Our study specifically focuses on the inter-county mobility pattern and its influence on the COVID-19 spread in the United States. To retrieve real-world mobility patterns, we utilize an integrated set of mobile device location data including over 100 million anonymous devices. We first investigate the nationwide temporal trend and spatial distribution of inter-county mobility. Then we zoom in on the epicenter of the U.S. outbreak, New York City, and evaluate the impacts of its outflow on other counties. Finally, we develop a "log-linear double-risk" model at the county level to quantify the influence of both "external risk" imported by inter-county mobility flows and the "internal risk" defined as the vulnerability of a county in terms of population with high-risk phenotypes. Our study enhances the situation awareness of inter-county mobility in the U.S. and can help improve non-pharmaceutical interventions for COVID-19.

preprint2020arXiv

Quarantine Fatigue: first-ever decrease in social distancing measures after the COVID-19 outbreak before reopening United States

By the emergence of the novel coronavirus disease (COVID-19) in Wuhan, China, and its rapid outbreak worldwide, the infectious illness has changed our everyday travel patterns. In this research, our team investigated the changes in the daily mobility pattern of people during the pandemic by utilizing an integrated data panel. To incorporate various aspects of human mobility, the team focused on the Social Distancing Index (SDI) which was calculated based on five basic mobility measures. The SDI patterns showed a plateau stage in the beginning of April that lasted for about two weeks. This phenomenon then followed by a universal decline of SDI, increased number of trips and reduction in percentage of people staying at home. We called the observation Quarantine Fatigue. The Rate of Change (ROC) method was employed to trace back the start date of quarantine fatigue which was indicated to be April 15th. Our analysis showed that despite the existence of state-to-state variations, most states started experiencing a quarantine fatigue phenomenon during the same period. This observation became more important by knowing that none of the states had officially announced the reopening until late April showing that people decided to loosen up their social distancing practices before the official reopening announcement. Moreover, our analysis indicated that official reopening led to a rapid decline in SDI, raising the concern of a second wave of outbreak. The synchronized trend among states also emphasizes the importance of a more nationwide decision-making attitude for the future as the condition of each state depends on the nationwide behavior.

preprint2020arXiv

Real-time Human Activity Recognition Using Conditionally Parametrized Convolutions on Mobile and Wearable Devices

Recently, deep learning has represented an important research trend in human activity recognition (HAR). In particular, deep convolutional neural networks (CNNs) have achieved state-of-the-art performance on various HAR datasets. For deep learning, improvements in performance have to heavily rely on increasing model size or capacity to scale to larger and larger datasets, which inevitably leads to the increase of operations. A high number of operations in deep leaning increases computational cost and is not suitable for real-time HAR using mobile and wearable sensors. Though shallow learning techniques often are lightweight, they could not achieve good performance. Therefore, deep learning methods that can balance the trade-off between accuracy and computation cost is highly needed, which to our knowledge has seldom been researched. In this paper, we for the first time propose a computation efficient CNN using conditionally parametrized convolution for real-time HAR on mobile and wearable devices. We evaluate the proposed method on four public benchmark HAR datasets consisting of WISDM dataset, PAMAP2 dataset, UNIMIB-SHAR dataset, and OPPORTUNITY dataset, achieving state-of-the-art accuracy without compromising computation cost. Various ablation experiments are performed to show how such a network with large capacity is clearly preferable to baseline while requiring a similar amount of operations. The method can be used as a drop-in replacement for the existing deep HAR architectures and easily deployed onto mobile and wearable devices for real-time HAR applications.

preprint2020arXiv

Review and Critical Analysis of Privacy-preserving Infection Tracking and Contact Tracing

The outbreak of viruses have necessitated contact tracing and infection tracking methods. Despite various efforts, there is currently no standard scheme for the tracing and tracking. Many nations of the world have therefore, developed their own ways where carriers of disease could be tracked and their contacts traced. These are generalized methods developed either in a distributed manner giving citizens control of their identity or in a centralised manner where a health authority gathers data on those who are carriers. This paper outlines some of the most significant approaches that have been established for contact tracing around the world. A comprehensive review on the key enabling methods used to realise the infrastructure around these infection tracking and contact tracing methods is also presented and recommendations are made for the most effective way to develop such a practice.

preprint2020arXiv

Search for New Hadronic Decays of $h_c$ and Observation of $h_c\rightarrow K^{+}K^{-}π^{+}π^{-}π^{0}$

Ten hadronic final states of the $h_c$ decays are investigated via the process $ψ(3686)\rightarrow π^0 h_c$, using a data sample of $(448.1 \pm 2.9) \times 10^6$ $ψ(3686)$ events collected with the BESIII detector. The decay channel $h_c\rightarrow K^{+}K^{-}π^{+}π^{-}π^{0}$ is observed for the first time with a significance of $6.0 σ$. The corresponding branching fraction is determined to be $\mathcal{B}(h_c\rightarrow K^{+}K^{-}π^{+}π^{-}π^{0}) =(3.3 \pm 0.6 \pm 0.6)\times 10^{-3}$ (the first uncertainty is statistical and the second systematical). Evidence for the decays $h_c\rightarrow π^{+} π^{-} π^{0} η$ and $h_c\rightarrow K^{0}_{S}K^{\pm}π^{\mp}π^{+}π^{-}$ is found with a significance of $3.6 σ$ and $3.8 σ$, respectively. The corresponding branching fractions (and upper limits) are obtained to be $\mathcal{B}(h_c\rightarrow π^{+} π^{-} π^{0} η) =(7.2 \pm 1.8 \pm 1.3)\times 10^{-3}$ $(< 1.8 \times 10^{-2})$ and $\mathcal{B}(h_c\rightarrow K^{0}_{S}K^{\pm}π^{\mp}π^{+}π^{-}) =(2.8 \pm 0.9 \pm 0.5)\times 10^{-3}$ $(<4.7\times 10^{-3})$. Upper limits on the branching fractions for the final states $h_c \rightarrow K^{+}K^{-}π^{0}$, $K^{+}K^{-}η$, $K^{+}K^{-}π^{+}π^{-}η$, $2(K^{+}K^{-})π^{0}$, $K^{+}K^{-}π^{0}η$, $K^{0}_{S}K^{\pm}π^{\mp}$, and $p\bar{p}π^{0}π^{0}$ are determined at a confidence level of 90\%.

preprint2020arXiv

Search for the decay $J/ψ\toγ+ \rm {invisible}$

We search for $J/ψ$ radiative decays into a weakly interacting neutral particle, namely an invisible particle, using the $J/ψ$ produced through the process $ψ(3686)\toπ^+π^-J/ψ$ in a data sample of $(448.1\pm2.9)\times 10^6$ $ψ(3686)$ decays collected by the BESIII detector at BEPCII. No significant signal is observed. Using a modified frequentist method, upper limits on the branching fractions are set under different assumptions of invisible particle masses up to 1.2 $\mathrm{\ Ge\kern -0.1em V}/c^2$. The upper limit corresponding to an invisible particle with zero mass is 7.0$\times 10^{-7}$ at the 90\% confidence level.

preprint2020arXiv

Search for the semileptonic decay $D^{0(+)}\to b_1(1235)^{-(0)} e^+ν_e$

Using $2.93~\mathrm{fb}^{-1}$ of $e^+e^-$ annihilation data collected at a center-of-mass energy $\sqrt{s}=3.773$ GeV with the BESIII detector operating at the BEPCII collider, we search for the semileptonic $D^{0(+)}$ decays into a $b_1(1235)^{-(0)}$ axial-vector meson for the first time. No significant signal is observed for either charge combination. The upper limits on the product branching fractions are ${\mathcal B}_{D^0\to b_1(1235)^- e^+ν_e}\cdot {\mathcal B}_{b_1(1235)^-\to ωπ^-}<1.12\times 10^{-4}$ and ${\mathcal B}_{D^+\to b_1(1235)^0 e^+ν_e}\cdot {\mathcal B}_{b_1(1235)^0\to ωπ^0}<1.75\times 10^{-4}$ at the 90\% confidence level.

preprint2020arXiv

Self-adaptive Re-weighted Adversarial Domain Adaptation

Existing adversarial domain adaptation methods mainly consider the marginal distribution and these methods may lead to either under transfer or negative transfer. To address this problem, we present a self-adaptive re-weighted adversarial domain adaptation approach, which tries to enhance domain alignment from the perspective of conditional distribution. In order to promote positive transfer and combat negative transfer, we reduce the weight of the adversarial loss for aligned features while increasing the adversarial force for those poorly aligned measured by the conditional entropy. Additionally, triplet loss leveraging source samples and pseudo-labeled target samples is employed on the confusing domain. Such metric loss ensures the distance of the intra-class sample pairs closer than the inter-class pairs to achieve the class-level alignment. In this way, the high accurate pseudolabeled target samples and semantic alignment can be captured simultaneously in the co-training process. Our method achieved low joint error of the ideal source and target hypothesis. The expected target error can then be upper bounded following Ben-David's theorem. Empirical evidence demonstrates that the proposed model outperforms state of the arts on standard domain adaptation datasets.

preprint2020arXiv

Single-Shot Two-Pronged Detector with Rectified IoU Loss

In the CNN based object detectors, feature pyramids are widely exploited to alleviate the problem of scale variation across object instances. These object detectors, which strengthen features via a top-down pathway and lateral connections, are mainly to enrich the semantic information of low-level features, but ignore the enhancement of high-level features. This can lead to an imbalance between different levels of features, in particular a serious lack of detailed information in the high-level features, which makes it difficult to get accurate bounding boxes. In this paper, we introduce a novel two-pronged transductive idea to explore the relationship among different layers in both backward and forward directions, which can enrich the semantic information of low-level features and detailed information of high-level features at the same time. Under the guidance of the two-pronged idea, we propose a Two-Pronged Network (TPNet) to achieve bidirectional transfer between high-level features and low-level features, which is useful for accurately detecting object at different scales. Furthermore, due to the distribution imbalance between the hard and easy samples in single-stage detectors, the gradient of localization loss is always dominated by the hard examples that have poor localization accuracy. This will enable the model to be biased toward the hard samples. So in our TPNet, an adaptive IoU based localization loss, named Rectified IoU (RIoU) loss, is proposed to rectify the gradients of each kind of samples. The Rectified IoU loss increases the gradients of examples with high IoU while suppressing the gradients of examples with low IoU, which can improve the overall localization accuracy of model. Extensive experiments demonstrate the superiority of our TPNet and RIoU loss.

preprint2020arXiv

SMAP: A Joint Dimensionality Reduction Scheme for Secure Multi-Party Visualization

Nowadays, as data becomes increasingly complex and distributed, data analyses often involve several related datasets that are stored on different servers and probably owned by different stakeholders. While there is an emerging need to provide these stakeholders with a full picture of their data under a global context, conventional visual analytical methods, such as dimensionality reduction, could expose data privacy when multi-party datasets are fused into a single site to build point-level relationships. In this paper, we reformulate the conventional t-SNE method from the single-site mode into a secure distributed infrastructure. We present a secure multi-party scheme for joint t-SNE computation, which can minimize the risk of data leakage. Aggregated visualization can be optionally employed to hide disclosure of point-level relationships. We build a prototype system based on our method, SMAP, to support the organization, computation, and exploration of secure joint embedding. We demonstrate the effectiveness of our approach with three case studies, one of which is based on the deployment of our system in real-world applications.

preprint2020arXiv

Study of BESIII Trigger Efficiencies with the 2018 $J/ψ$ Data

Using a dedicated data sample taken in 2018 on the $J/ψ$ peak, we perform a detailed study of the trigger efficiencies of the BESIII detector. The efficiencies are determined from three representative physics processes, namely Bhabha-scattering, dimuon production and generic hadronic events with charged particles. The combined efficiency of all active triggers approaches $100\%$ in most cases with uncertainties small enough as not to affect most physics analyses.

preprint2020arXiv

Study of open-charm decays and radiative transitions of the X(3872)

The processes $X(3872)\to D^{*0}\bar{D^{0}}+c.c.,~γJ/ψ,~γψ(2S),$ and $γD^{+}D^{-}$ are searched for in a $9.0~\rm fb^{-1}$ data sample collected at center-of-mass energies between $4.178$ and $4.278$ GeV with the BESIII detector. We observe $X(3872)\to D^{*0}\bar{D^{0}}+c.c.$ and find evidence for $X(3872)\toγJ/ψ$ with statistical significances of $7.4σ$ and $3.5σ$, respectively. No evident signals for $X(3872)\toγψ(2S)$ and $γD^{+}D^{-}$ are found, and upper limit on the relative branching ratio $R_{γψ} \equiv\frac{\mathcal{B}(X(3872)\toγψ(2S))}{\mathcal{B}(X(3872)\toγJ/ψ)}<0.59$ is set at 90$\%$ confidence level. Measurements of branching ratios relative to decay $X(3872)\toπ^+π^- J/ψ$ are also reported for decays $X(3872)\to D^{*0}\bar{D^{0}}+c.c., ~γψ(2S),~γJ/ψ$, $γD^{+}D^{-}$, as well as the non-$D^{*0}\bar{D}^{0}$ three-body decays $π^0 D^{0}\bar{D}^{0}$ and $γD^{0}\bar{D}^{0}$.

preprint2020arXiv

Subadditivity of Kodaira dimension does not hold in positive characteristic

Over any algebraically closed field of positive characteristic, we construct examples of fibrations violating subadditivity of Kodaira dimension.

preprint2020arXiv

Suppress and Balance: A Simple Gated Network for Salient Object Detection

Most salient object detection approaches use U-Net or feature pyramid networks (FPN) as their basic structures. These methods ignore two key problems when the encoder exchanges information with the decoder: one is the lack of interference control between them, the other is without considering the disparity of the contributions of different encoder blocks. In this work, we propose a simple gated network (GateNet) to solve both issues at once. With the help of multilevel gate units, the valuable context information from the encoder can be optimally transmitted to the decoder. We design a novel gated dual branch structure to build the cooperation among different levels of features and improve the discriminability of the whole network. Through the dual branch design, more details of the saliency map can be further restored. In addition, we adopt the atrous spatial pyramid pooling based on the proposed "Fold" operation (Fold-ASPP) to accurately localize salient objects of various scales. Extensive experiments on five challenging datasets demonstrate that the proposed model performs favorably against most state-of-the-art methods under different evaluation metrics.

preprint2020arXiv

Tasks Integrated Networks: Joint Detection and Retrieval for Image Search

The traditional object retrieval task aims to learn a discriminative feature representation with intra-similarity and inter-dissimilarity, which supposes that the objects in an image are manually or automatically pre-cropped exactly. However, in many real-world searching scenarios (e.g., video surveillance), the objects (e.g., persons, vehicles, etc.) are seldom accurately detected or annotated. Therefore, object-level retrieval becomes intractable without bounding-box annotation, which leads to a new but challenging topic, i.e. image-level search. In this paper, to address the image search issue, we first introduce an end-to-end Integrated Net (I-Net), which has three merits: 1) A Siamese architecture and an on-line pairing strategy for similar and dissimilar objects in the given images are designed. 2) A novel on-line pairing (OLP) loss is introduced with a dynamic feature dictionary, which alleviates the multi-task training stagnation problem, by automatically generating a number of negative pairs to restrict the positives. 3) A hard example priority (HEP) based softmax loss is proposed to improve the robustness of classification task by selecting hard categories. With the philosophy of divide and conquer, we further propose an improved I-Net, called DC-I-Net, which makes two new contributions: 1) two modules are tailored to handle different tasks separately in the integrated framework, such that the task specification is guaranteed. 2) A class-center guided HEP loss (C2HEP) by exploiting the stored class centers is proposed, such that the intra-similarity and inter-dissimilarity can be captured for ultimate retrieval. Extensive experiments on famous image-level search oriented benchmark datasets demonstrate that the proposed DC-I-Net outperforms the state-of-the-art tasks-integrated and tasks-separated image search models.

preprint2020arXiv

Topological one-way large-area waveguide states in magnetic photonic crystals

We have theoretically and experimentally achieved large-area one-way transport by using heterostructures consisting of a domain of an ordinary photonic crystal (PC) sandwiched between two domains of magnetic PCs. The non-magnetized domain carries two orthogonal one-way waveguide states which have amplitude uniformly distributed over a large-area. These two waveguide states support unidirectional transport even though the medium of propagation is not magnetized. We show both experimentally and numerically that such one-way waveguide states can be utilized to abruptly narrow the beam width of an extended state to concentrate energy. Such extended waveguide modes are robust to different kinds of defects, such as voids and PEC barriers. They are also immune to the Anderson type localization when large randomness is introduced.

preprint2020arXiv

Tuning Band Alignment and Optical Properties of 2D van der Waals Heterostructure via Ferroelectric Polarization Switching

Favourable band alignment and excellent visible light response are vital for photochemical water splitting. In this work, we have theoretically investigated how ferroelectric polarization and its reversibility in direction can be utilized to modulate the band alignment and optical absorption properties. For this objective, 2D van der Waals heterostructures (HTSs) are constructed by interfacing monolayer MoS2 with ferroelectric In2Se3. We find the switch of polarization direction has dramatically changed the band alignment, thus facilitating different type of reactions. In In2Se3/MoS2/In2Se3 heterostructures, one polarization direction supports hydrogen evolution reaction and another polarization direction can favour oxygen evolution reaction. These can be used to create tuneable photocatalyst materials where water reduction reactions can be selectively controlled by polarization switching. The modulation of band alignment is attributed to the shift of reaction potential caused by spontaneous polarization. Additionally, the formed type-II van der Waals HTSs also significantly improve charge separation and enhance the optical absorption in the visible and infrared regions. Our results pave a way in the design of van der Waals HTSs for water splitting using ferroelectric materials.

preprint2020arXiv

Weakness Analysis of Cyberspace Configuration Based on Reinforcement Learning

In this work, we present a learning-based approach to analysis cyberspace configuration. Unlike prior methods, our approach has the ability to learn from past experience and improve over time. In particular, as we train over a greater number of agents as attackers, our method becomes better at rapidly finding attack paths for previously hidden paths, especially in multiple domain cyberspace. To achieve these results, we pose finding attack paths as a Reinforcement Learning (RL) problem and train an agent to find multiple domain attack paths. To enable our RL policy to find more hidden attack paths, we ground representation introduction an multiple domain action select module in RL. By designing a simulated cyberspace experimental environment to verify our method. Our objective is to find more hidden attack paths, to analysis the weakness of cyberspace configuration. The experimental results show that our method can find more hidden multiple domain attack paths than existing baselines methods.

preprint2020arXiv

What makes an explosion happen?

The presence of the nonbonding XH tension constrains and the presence of the anti or super hydrogen bond fosters the explosion in aqueous alkali and molten alkali halides; the combination of the coupled hydrogen bond and the repulsive anti or super hydrogen bond not only stabilzes the structure but also stores energy of the energetic molecular assemblies by shortening all covalent bonds.

Institution

Affiliation not imported yet

This author record came from a source that does not expose affiliation metadata. Once the author claims the profile or we enrich the record from another provider, this section will link to the concrete institution.

Source provenance

Where this author record came from

arxivconfidence 95%

external id: arxiv:2605.04874:author:3:lei-zhang

Imported May 20, 2026Synced May 21, 2026

arxivconfidence 95%

external id: arxiv:2604.26342:author:1:lei-zhang