Source author record

Sheng Wang

Sheng Wang appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Catalog footprint

What is connected

68works

35topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

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

preprint2026arXiv

AutoResearchClaw: Self-Reinforcing Autonomous Research with Human-AI Collaboration

Automating scientific discovery requires more than generating papers from ideas. Real research is iterative: hypotheses are challenged from multiple perspectives, experiments fail and inform the next attempt, and lessons accumulate across cycles. Existing autonomous research systems often model this process as a linear pipeline: they rely on single-agent reasoning, stop when execution fails, and do not carry experience across runs. We present AutoResearchClaw, a multi-agent autonomous research pipeline built on five mechanisms: structured multi-agent debate for hypothesis generation and result analysis, a self-healing executor with a \textsc{Pivot}/\textsc{Refine} decision loop that transforms failures into information, verifiable result reporting that prevents fabricated numbers and hallucinated citations, human-in-the-loop collaboration with seven intervention modes spanning full autonomy to step-by-step oversight, and cross-run evolution that converts past mistakes into future safeguards. On ARC-Bench, a 25-topic experiment-stage benchmark, AutoResearchClaw outperforms AI Scientist v2 by 54.7%. A human-in-the-loop ablation across seven intervention modes reveals that precise, targeted collaboration at high-leverage decision points consistently outperforms both full autonomy and exhaustive step-by-step oversight. We position AutoResearchClaw as a research amplifier that augments rather than replaces human scientific judgment. Code is available at https://github.com/aiming-lab/AutoResearchClaw.

preprint2026arXiv

Hypergraph Pattern Machine: Compositional Tokenization for Higher-Order Interactions

Hypergraphs model higher-order relations that drive real-world decisions, from drug prescriptions to recommendations. A central structural signal in such data, beyond what pairwise relations can express, is interaction compositionality: whether a higher-order relation is compositional, emergent, or inhibitory with respect to its observed or unobserved sets. In polypharmacy, the regime decides whether a drug should be dropped, kept, or excluded: a compositional drug triple can be safely simplified, an emergent triple requires all drugs jointly, and an inhibitory triple flags a drug that disrupts an existing interaction. However, existing hypergraph learning methods, which merely propagate messages over observed hyperedges, leave this compositional signal unmodeled, allowing dangerous drug combinations to slip through and be misclassified. To this end, we propose the Hypergraph Pattern Machine (HGPM), shifting the paradigm from message passing to learning the compositional pattern of subsets. It tokenizes compositional subsets, organizes them in an inclusion DAG, and trains an inclusion-aware Transformer under masked reconstruction. On ten hypergraph benchmarks, HGPM matches or exceeds state-of-the-art methods. Notably, in a real adverse-event prediction case, HGPM correctly identifies the drug addition that inhibits the side effect among feature-identical candidates, a discrimination existing methods cannot make. The code and data are in https://github.com/KryieZhao/HGPM.git.

preprint2026arXiv

MicroDiffuse3D: A Foundation Model for 3D Microscopy Imaging Restoration

Chemical imaging enables label-free visualization of cells, tissues and living systems while providing direct biochemical information that is difficult to obtain with conventional fluorescence microscopy. Despite its promise in applications ranging from intraoperative diagnosis to drug-response analysis, its broader use remains limited by slow data acquisition, particularly for three-dimensional imaging. Here we present MicroDiffuse3D, a pretrained foundation model for 3D microscopy image restoration that recovers high-quality volumetric structure from degraded low-resolution measurements acquired at substantially higher throughput. We evaluated MicroDiffuse3D across three challenging restoration settings, including 3D super-resolution under 16-fold volumetric sparsity, joint degradation in resolution and noise, and 3D denoising in the low signal-to-noise ratio (SNR) regime, where the model delivered clear gains over strong baselines. Under the sparse 3D super-resolution setting, MicroDiffuse3D produced clearer continuity across depth with fewer artifacts and improved segmentation quality by 10.58% and line-profile concordance by 15.59%. Together, our results establish pretrained 3D restoration as a broadly applicable strategy for overcoming the throughput and SNR limitations in volumetric chemical imaging, enabling high-resolution analysis at scales and speeds that were previously difficult to achieve.

preprint2026arXiv

PRISM: Fast Online LLM Serving via Scheduling-Memory Co-design

Modern online large language model (LLM) services, such as Retrieval-Augmented Generation (RAG) and agent systems, increasingly expose two prominent characteristics: prompt segmentation (e.g., system instructions, retrieved passages, tool outputs) and hotspot skew, where a small set of these segments recurs frequently across user requests. Failing to jointly exploit these patterns could lead to repeated prefill of hot segments and prolonged TTFT, undermining both throughput and user-perceived responsiveness. However, existing work tackles these patterns independently: KV-cache management mainly exploits segment reuse while scheduling reorders requests to improve cache locality, yet neither aligns request admission with KV-cache retention. To address this gap, we first analyze how scheduling and KV-cache management jointly affect TTFT. Guided by this, we present PRISM (Prefix Reuse Optimization Integrated Scheduling and Memory), which co-designs a query-aware scheduler (QAS) with a demand-aware radix tree (DART) to align request admission with exact-prefix KV retention. Our evaluation results show that, versus the strongest baseline, PRISM reduces average per-QPS P99 TTFT by 23.3\% and 37.1\% while increasing exact-prefix KV-cache hit rate by 5.9 and 12.2 percentage points on 4B and 13B models, respectively.

preprint2026arXiv

Relay Buffer Independent Communication over Pooled HBM for Efficient MoE Inference on Ascend

Mixture-of-Experts (MoE) inference requires large-scale token exchange across devices, making dispatch and combine major bottlenecks in both prefill and decode. Beyond network transfer, routing-driven layout transformation, temporary relay, and output restoration can add substantial overhead. Existing MoE communication paths are often buffer-centric, using explicit inter-process relay and reordering buffers around collective transfer. This report presents a relay-buffer-free communication design for MoE inference acceleration on Ascend systems. The design reorganizes dispatch and combine around direct placement into destination expert windows and direct reading from remote expert windows. Built on globally pooled high-bandwidth memory and symmetric-memory allocation, it removes most intermediate relay and reordering buffers while retaining only lightweight control state, including counts, offsets, and synchronization metadata. We instantiate the design as two schedules for the main phases of MoE inference: a prefill schedule with richer planning state for throughput-oriented execution, and a compact decode schedule for latency-sensitive execution. Experiments on Ascend-based MoE workloads show reduced dispatch and combine latency in both settings. At the serving level, the implementation improves time to first token (TTFT), preserves competitive time per output token (TPOT), and enlarges the feasible scheduling space under practical latency constraints. These results indicate that, on platforms with globally addressable device memory, reducing intermediate buffering and output restoration around expert execution is an effective direction for accelerating MoE inference.

preprint2024arXiv

Long time behaviors for the inhomogeneous NLS with a potential in $\mathbb{R}^3$

In this article, we aim to study the scattering of the solution to the focusing inhomogeneous nonlinear Schrödinger equation with a potential of form \begin{align*} i\partial_t u+Δu- Vu=-|x|^{-b}|u|^{p-1}u \end{align*} in the energy space $H^1(\R^3)$. We prove a scattering criterion, and then we use it together with Morawetz estimate to show the scattering theory, which generalizes the results of Dinh \cite{DD} to the non-radial symmetric case.

preprint2022arXiv

Adaptive Transfer Learning for Plant Phenotyping

Plant phenotyping (Guo et al. 2021; Pieruschka et al. 2019) focuses on studying the diverse traits of plants related to the plants' growth. To be more specific, by accurately measuring the plant's anatomical, ontogenetical, physiological and biochemical properties, it allows identifying the crucial factors of plants' growth in different environments. One commonly used approach is to predict the plant's traits using hyperspectral reflectance (Yendrek et al. 2017; Wang et al. 2021). However, the data distributions of the hyperspectral reflectance data in plant phenotyping might vary in different environments for different plants. That is, it would be computationally expansive to learn the machine learning models separately for one plant in different environments. To solve this problem, we focus on studying the knowledge transferability of modern machine learning models in plant phenotyping. More specifically, this work aims to answer the following questions. (1) How is the performance of conventional machine learning models, e.g., partial least squares regression (PLSR), Gaussian process regression (GPR) and multi-layer perceptron (MLP), affected by the number of annotated samples for plant phenotyping? (2) Whether could the neural network based transfer learning models improve the performance of plant phenotyping? (3) Could the neural network based transfer learning be improved by using infinite-width hidden layers for plant phenotyping?

preprint2022arXiv

Brain-Aware Replacements for Supervised Contrastive Learning in Detection of Alzheimer's Disease

We propose a novel framework for Alzheimer's disease (AD) detection using brain MRIs. The framework starts with a data augmentation method called Brain-Aware Replacements (BAR), which leverages a standard brain parcellation to replace medically-relevant 3D brain regions in an anchor MRI from a randomly picked MRI to create synthetic samples. Ground truth "hard" labels are also linearly mixed depending on the replacement ratio in order to create "soft" labels. BAR produces a great variety of realistic-looking synthetic MRIs with higher local variability compared to other mix-based methods, such as CutMix. On top of BAR, we propose using a soft-label-capable supervised contrastive loss, aiming to learn the relative similarity of representations that reflect how mixed are the synthetic MRIs using our soft labels. This way, we do not fully exhaust the entropic capacity of our hard labels, since we only use them to create soft labels and synthetic MRIs through BAR. We show that a model pre-trained using our framework can be further fine-tuned with a cross-entropy loss using the hard labels that were used to create the synthetic samples. We validated the performance of our framework in a binary AD detection task against both from-scratch supervised training and state-of-the-art self-supervised training plus fine-tuning approaches. Then we evaluated BAR's individual performance compared to another mix-based method CutMix by integrating it within our framework. We show that our framework yields superior results in both precision and recall for the AD detection task.

preprint2022arXiv

Estimating risks of option books using neural-SDE market models

In this paper, we examine the capacity of an arbitrage-free neural-SDE market model to produce realistic scenarios for the joint dynamics of multiple European options on a single underlying. We subsequently demonstrate its use as a risk simulation engine for option portfolios. Through backtesting analysis, we show that our models are more computationally efficient and accurate for evaluating the Value-at-Risk (VaR) of option portfolios, with better coverage performance and less procyclicality than standard filtered historical simulation approaches.

preprint2022arXiv

Eye-gaze-guided Vision Transformer for Rectifying Shortcut Learning

Learning harmful shortcuts such as spurious correlations and biases prevents deep neural networks from learning the meaningful and useful representations, thus jeopardizing the generalizability and interpretability of the learned representation. The situation becomes even more serious in medical imaging, where the clinical data (e.g., MR images with pathology) are limited and scarce while the reliability, generalizability and transparency of the learned model are highly required. To address this problem, we propose to infuse human experts' intelligence and domain knowledge into the training of deep neural networks. The core idea is that we infuse the visual attention information from expert radiologists to proactively guide the deep model to focus on regions with potential pathology and avoid being trapped in learning harmful shortcuts. To do so, we propose a novel eye-gaze-guided vision transformer (EG-ViT) for diagnosis with limited medical image data. We mask the input image patches that are out of the radiologists' interest and add an additional residual connection in the last encoder layer of EG-ViT to maintain the correlations of all patches. The experiments on two public datasets of INbreast and SIIM-ACR demonstrate our EG-ViT model can effectively learn/transfer experts' domain knowledge and achieve much better performance than baselines. Meanwhile, it successfully rectifies the harmful shortcut learning and significantly improves the EG-ViT model's interpretability. In general, EG-ViT takes the advantages of both human expert's prior knowledge and the power of deep neural networks. This work opens new avenues for advancing current artificial intelligence paradigms by infusing human intelligence.

preprint2022arXiv

Follow My Eye: Using Gaze to Supervise Computer-Aided Diagnosis

When deep neural network (DNN) was first introduced to the medical image analysis community, researchers were impressed by its performance. However, it is evident now that a large number of manually labeled data is often a must to train a properly functioning DNN. This demand for supervision data and labels is a major bottleneck in current medical image analysis, since collecting a large number of annotations from experienced experts can be time-consuming and expensive. In this paper, we demonstrate that the eye movement of radiologists reading medical images can be a new form of supervision to train the DNN-based computer-aided diagnosis (CAD) system. Particularly, we record the tracks of the radiologists' gaze when they are reading images. The gaze information is processed and then used to supervise the DNN's attention via an Attention Consistency module. To the best of our knowledge, the above pipeline is among the earliest efforts to leverage expert eye movement for deep-learning-based CAD. We have conducted extensive experiments on knee X-ray images for osteoarthritis assessment. The results show that our method can achieve considerable improvement in diagnosis performance, with the help of gaze supervision.

preprint2022arXiv

Graph-in-Graph Network for Automatic Gene Ontology Description Generation

Gene Ontology (GO) is the primary gene function knowledge base that enables computational tasks in biomedicine. The basic element of GO is a term, which includes a set of genes with the same function. Existing research efforts of GO mainly focus on predicting gene term associations. Other tasks, such as generating descriptions of new terms, are rarely pursued. In this paper, we propose a novel task: GO term description generation. This task aims to automatically generate a sentence that describes the function of a GO term belonging to one of the three categories, i.e., molecular function, biological process, and cellular component. To address this task, we propose a Graph-in-Graph network that can efficiently leverage the structural information of GO. The proposed network introduces a two-layer graph: the first layer is a graph of GO terms where each node is also a graph (gene graph). Such a Graph-in-Graph network can derive the biological functions of GO terms and generate proper descriptions. To validate the effectiveness of the proposed network, we build three large-scale benchmark datasets. By incorporating the proposed Graph-in-Graph network, the performances of seven different sequence-to-sequence models can be substantially boosted across all evaluation metrics, with up to 34.7%, 14.5%, and 39.1% relative improvements in BLEU, ROUGE-L, and METEOR, respectively.

preprint2022arXiv

Hedging option books using neural-SDE market models

We study the capability of arbitrage-free neural-SDE market models to yield effective strategies for hedging options. In particular, we derive sensitivity-based and minimum-variance-based hedging strategies using these models and examine their performance when applied to various option portfolios using real-world data. Through backtesting analysis over typical and stressed market periods, we show that neural-SDE market models achieve lower hedging errors than Black--Scholes delta and delta-vega hedging consistently over time, and are less sensitive to the tenor choice of hedging instruments. In addition, hedging using market models leads to similar performance to hedging using Heston models, while the former tends to be more robust during stressed market periods.

preprint2022arXiv

Joint Progressive and Coarse-to-fine Registration of Brain MRI via Deformation Field Integration and Non-Rigid Feature Fusion

Registration of brain MRI images requires to solve a deformation field, which is extremely difficult in aligning intricate brain tissues, e.g., subcortical nuclei, etc. Existing efforts resort to decomposing the target deformation field into intermediate sub-fields with either tiny motions, i.e., progressive registration stage by stage, or lower resolutions, i.e., coarse-to-fine estimation of the full-size deformation field. In this paper, we argue that those efforts are not mutually exclusive, and propose a unified framework for robust brain MRI registration in both progressive and coarse-to-fine manners simultaneously. Specifically, building on a dual-encoder U-Net, the fixed-moving MRI pair is encoded and decoded into multi-scale deformation sub-fields from coarse to fine. Each decoding block contains two proposed novel modules: i) in Deformation Field Integration (DFI), a single integrated sub-field is calculated, warping by which is equivalent to warping progressively by sub-fields from all previous decoding blocks, and ii) in Non-rigid Feature Fusion (NFF), features of the fixed-moving pair are aligned by DFI-integrated sub-field, and then fused to predict a finer sub-field. Leveraging both DFI and NFF, the target deformation field is factorized into multi-scale sub-fields, where the coarser fields alleviate the estimate of a finer one and the finer field learns to make up those misalignments insolvable by previous coarser ones. The extensive and comprehensive experimental results on both private and public datasets demonstrate a superior registration performance of brain MRI images over progressive registration only and coarse-to-fine estimation only, with an increase by at most 8% in the average Dice.

preprint2022arXiv

Knee Cartilage Defect Assessment by Graph Representation and Surface Convolution

Knee osteoarthritis (OA) is the most common osteoarthritis and a leading cause of disability. Cartilage defects are regarded as major manifestations of knee OA, which are visible by magnetic resonance imaging (MRI). Thus early detection and assessment for knee cartilage defects are important for protecting patients from knee OA. In this way, many attempts have been made on knee cartilage defect assessment by applying convolutional neural networks (CNNs) to knee MRI. However, the physiologic characteristics of the cartilage may hinder such efforts: the cartilage is a thin curved layer, implying that only a small portion of voxels in knee MRI can contribute to the cartilage defect assessment; heterogeneous scanning protocols further challenge the feasibility of the CNNs in clinical practice; the CNN-based knee cartilage evaluation results lack interpretability. To address these challenges, we model the cartilages structure and appearance from knee MRI into a graph representation, which is capable of handling highly diverse clinical data. Then, guided by the cartilage graph representation, we design a non-Euclidean deep learning network with the self-attention mechanism, to extract cartilage features in the local and global, and to derive the final assessment with a visualized result. Our comprehensive experiments show that the proposed method yields superior performance in knee cartilage defect assessment, plus its convenient 3D visualization for interpretability.

preprint2022arXiv

Non-degeneracy of Stochastic Line Integrals

We derive quantitative criteria for the existence of density for stochastic line integrals and iterated line integrals along solutions of hypoelliptic differential equations driven by fractional Brownian motion. As an application, we also study the signature uniqueness problem for these rough differential equations.

preprint2022arXiv

ProTranslator: zero-shot protein function prediction using textual description

Accurately finding proteins and genes that have a certain function is the prerequisite for a broad range of biomedical applications. Despite the encouraging progress of existing computational approaches in protein function prediction, it remains challenging to annotate proteins to a novel function that is not collected in the Gene Ontology and does not have any annotated proteins. This limitation, a side effect from the widely-used multi-label classification problem setting of protein function prediction, hampers the progress of studying new pathways and biological processes, and further slows down research in various biomedical areas. Here, we tackle this problem by annotating proteins to a function only based on its textual description so that we do not need to know any associated proteins for this function. The key idea of our method ProTranslator is to redefine protein function prediction as a machine translation problem, which translates the description word sequence of a function to the amino acid sequence of a protein. We can then transfer annotations from functions that have similar textual description to annotate a novel function. We observed substantial improvement in annotating novel functions and sparsely annotated functions on CAFA3, SwissProt and GOA datasets. We further demonstrated how our method accurately predicted gene members for a given pathway in Reactome, KEGG and MSigDB only based on the pathway description. Finally, we showed how ProTranslator enabled us to generate the textual description instead of the function label for a set of proteins, providing a new scheme for protein function prediction. We envision ProTranslator will give rise to a protein function "search engine" that returns a list of proteins based on the free text queried by the user.

preprint2022arXiv

Scattering For three waves Nonlinear Schrödinger System with mass-resonance in 5D

In this paper, we study the dynamics behavior of the NLS system with three waves interaction in the energy space $H^1(\mathbb{R}^5) \times H^1(\mathbb{R}^5)\times H^1(\mathbb{R}^5) $. Inspired by B. Dodson and J. Murphy in \cite{Dodson2018}, we establish an interaction Morawetz estimate for the NLS system, together with the criterion which proved by Tao-Dodson--Murphy we can get the scattering under the ground state in energy space with mass-resonance. Under the radial assumption, we can remove the mass-resonance condition.

preprint2022arXiv

Spin-charge separation in a 1D Fermi gas with tunable interactions

Ultracold atoms confined to periodic potentials have proven to be a powerful tool for quantum simulation of complex many-body systems. We confine fermions to one-dimension to realize the Tomonaga-Luttinger liquid model describing the highly collective nature of their low-energy excitations. We use Bragg spectroscopy to directly excite either the spin or charge wave for various strength of repulsive interaction. We observe that the velocity of the spin and charge excitations shift in opposite directions with increasing interaction, a hallmark of spin-charge separation. The excitation spectra are in quantitative agreement with the Tomonaga-Luttinger liquid theory, and furthermore, we find that the spin excitations become dispersive at large interaction, signaling the onset of the nonlinear Luttinger liquid regime.

preprint2022arXiv

TBI-GAN: An Adversarial Learning Approach for Data Synthesis on Traumatic Brain Segmentation

Brain network analysis for traumatic brain injury (TBI) patients is critical for its consciousness level assessment and prognosis evaluation, which requires the segmentation of certain consciousness-related brain regions. However, it is difficult to construct a TBI segmentation model as manually annotated MR scans of TBI patients are hard to collect. Data augmentation techniques can be applied to alleviate the issue of data scarcity. However, conventional data augmentation strategies such as spatial and intensity transformation are unable to mimic the deformation and lesions in traumatic brains, which limits the performance of the subsequent segmentation task. To address these issues, we propose a novel medical image inpainting model named TBI-GAN to synthesize TBI MR scans with paired brain label maps. The main strength of our TBI-GAN method is that it can generate TBI images and corresponding label maps simultaneously, which has not been achieved in the previous inpainting methods for medical images. We first generate the inpainted image under the guidance of edge information following a coarse-to-fine manner, and then the synthesized intensity image is used as the prior for label inpainting. Furthermore, we introduce a registration-based template augmentation pipeline to increase the diversity of the synthesized image pairs and enhance the capacity of data augmentation. Experimental results show that the proposed TBI-GAN method can produce sufficient synthesized TBI images with high quality and valid label maps, which can greatly improve the 2D and 3D traumatic brain segmentation performance compared with the alternatives.

preprint2022arXiv

Towards Accurate Active Camera Localization

In this work, we tackle the problem of active camera localization, which controls the camera movements actively to achieve an accurate camera pose. The past solutions are mostly based on Markov Localization, which reduces the position-wise camera uncertainty for localization. These approaches localize the camera in the discrete pose space and are agnostic to the localization-driven scene property, which restricts the camera pose accuracy in the coarse scale. We propose to overcome these limitations via a novel active camera localization algorithm, composed of a passive and an active localization module. The former optimizes the camera pose in the continuous pose space by establishing point-wise camera-world correspondences. The latter explicitly models the scene and camera uncertainty components to plan the right path for accurate camera pose estimation. We validate our algorithm on the challenging localization scenarios from both synthetic and scanned real-world indoor scenes. Experimental results demonstrate that our algorithm outperforms both the state-of-the-art Markov Localization based approach and other compared approaches on the fine-scale camera pose accuracy. Code and data are released at https://github.com/qhFang/AccurateACL.

preprint2021arXiv

Combined Trojan Y Chromosome Strategy and Sterile Insect Technique to Eliminate Mosquitoes: Modelling and Analysis

Sterile insect technique has been successfully applied in the control of agricultural pests, however, it has a limited ability to control mosquitoes. A promising alternative approach is Trojan Y Chromosome strategy, which works by manipulating the sex ratio of a population through the introduction of feminized Y Y supermales that guarantee male offspring. A combined Trojan Y chromosome strategy and sterile insect technique (TYC-SIT) strategy is modeled with ordinary differential equations that allow the kinetics of the female population decline of mosquitoes to be evaluated under identical modeling conditions. The dynamical analysis leads to results on both local and global stabilities of this combined model. Optimal control analysis is also implemented to investigate the optimal mechanisms for extinction of mosquitoes. In particular, the numerical results affirm that the combined TYC-SIT enables near elimination of mosquitoes. The conclusion has great significance for pest controls.

preprint2021arXiv

Field-effect at electrical contacts to two-dimensional materials

The inferior electrical contact to two-dimensional (2D) materials is a critical challenge for their application in post-silicon very large-scale integrated circuits. Electrical contacts were generally related to their resistive effect, quantified as contact resistance. With a systematic investigation, this work demonstrates a capacitive metal-insulator-semiconductor (MIS) field-effect at the electrical contacts to 2D materials: the field-effect depletes or accumulates charge carriers, redistributes the voltage potential, and give rise to abnormal current saturation and nonlinearity. On the one hand, the current saturation hinders the devices' driving ability, which can be eliminated with carefully engineered contact configurations. On the other hand, by introducing the nonlinearity to monolithic analog artificial neural network circuits, the circuits' perception ability can be significantly enhanced, as evidenced using a COVID-19 critical illness prediction model. This work provides a comprehension of the field-effect at the electrical contacts to 2D materials, which is fundamental to the design, simulation, and fabrication of electronics based on 2D material.

preprint2021arXiv

Radiation-tolerant high-entropy alloys via interstitial-solute-induced chemical heterogeneities

High-entropy alloys (HEAs) composed of multiple principal elements have been shown to offer improved radiation resistance over their elemental or dilute-solution counterparts. Using NiCoFeCrMn HEA as a model, here we introduce carbon and nitrogen interstitial alloying elements to impart chemical heterogeneities in the form of the local chemical order (LCO) and associated compositional variations. Density functional theory simulations predict chemical short-range order (CSRO) (nearest neighbors and the next couple of atomic shells) surrounding C and N, due to the chemical affinity of C with (Co, Fe) and N with (Cr, Mn). Atomic-resolution chemical mapping of the elemental distribution confirms marked compositional variations well beyond statistical fluctuations. Ni+ irradiation experiments at elevated temperatures demonstrate a remarkable reduction in void swelling by at least one order of magnitude compared to the base HEA without C and N alloying. The underlying mechanism is that the interstitial-solute-induced chemical heterogeneities roughen the lattice as well as the energy landscape, impeding the movements of, and constraining the path lanes for, the normally fast-moving self-interstitials and their clusters. The irradiation-produced interstitials and vacancies therefore recombine more readily, delaying void formation. Our findings thus open a promising avenue towards highly radiation-tolerant alloys.

preprint2020arXiv

Analysis of Indexing Structures for Immutable Data

In emerging applications such as blockchains and collaborative data analytics, there are strong demands for data immutability, multi-version accesses, and tamper-evident controls. This leads to three new index structures for immutable data, namely Merkle Patricia Trie (MPT), Merkle Bucket Tree (MBT), and Pattern-Oriented-Split Tree (POS-Tree). Although these structures have been adopted in real applications, there is no systematic evaluation of their pros and cons in the literature. This makes it difficult for practitioners to choose the right index structure for their applications, as there is only a limited understanding of the characteristics of each index. To alleviate the above deficiency, we present a comprehensive analysis of the existing index structures for immutable data, evaluating both their asymptotic and empirical performance. Specifically, we show that MPT, MBT, and POS-Tree are all instances of a recently proposed framework, dubbed \my{Structurally Invariant and Reusable Indexes (SIRI)}. We propose to evaluate the SIRI instances based on five essential metrics: their efficiency for four index operations (i.e., lookup, update, comparison, and merge), as well as their \my{deduplication ratios} (i.e., the size of the index with deduplication over the size without deduplication). We establish the worst-case guarantees of each index in terms of these five metrics, and we experimentally evaluate all indexes in a large variety of settings. Based on our theoretical and empirical analysis, we conclude that POS-Tree is a favorable choice for indexing immutable data.

preprint2020arXiv

ForkBase: Immutable, Tamper-evident Storage Substrate for Branchable Applications

Data collaboration activities typically require systematic or protocol-based coordination to be scalable. Git, an effective enabler for collaborative coding, has been attested for its success in countless projects around the world. Hence, applying the Git philosophy to general data collaboration beyond coding is motivating. We call it Git for data. However, the original Git design handles data at the file granule, which is considered too coarse-grained for many database applications. We argue that Git for data should be co-designed with database systems. To this end, we developed ForkBase to make Git for data practical. ForkBase is a distributed, immutable storage system designed for data version management and data collaborative operation. In this demonstration, we show how ForkBase can greatly facilitate collaborative data management and how its novel data deduplication technique can improve storage efficiency for archiving massive data versions.

preprint2020arXiv

Label-Driven Reconstruction for Domain Adaptation in Semantic Segmentation

Unsupervised domain adaptation enables to alleviate the need for pixel-wise annotation in the semantic segmentation. One of the most common strategies is to translate images from the source domain to the target domain and then align their marginal distributions in the feature space using adversarial learning. However, source-to-target translation enlarges the bias in translated images and introduces extra computations, owing to the dominant data size of the source domain. Furthermore, consistency of the joint distribution in source and target domains cannot be guaranteed through global feature alignment. Here, we present an innovative framework, designed to mitigate the image translation bias and align cross-domain features with the same category. This is achieved by 1) performing the target-to-source translation and 2) reconstructing both source and target images from their predicted labels. Extensive experiments on adapting from synthetic to real urban scene understanding demonstrate that our framework competes favorably against existing state-of-the-art methods.

preprint2020arXiv

mr2NST: Multi-Resolution and Multi-Reference Neural Style Transfer for Mammography

Computer-aided diagnosis with deep learning techniques has been shown to be helpful for the diagnosis of the mammography in many clinical studies. However, the image styles of different vendors are very distinctive, and there may exist domain gap among different vendors that could potentially compromise the universal applicability of one deep learning model. In this study, we explicitly address style variety issue with the proposed multi-resolution and multi-reference neural style transfer (mr2NST) network. The mr2NST can normalize the styles from different vendors to the same style baseline with very high resolution. We illustrate that the image quality of the transferred images is comparable to the quality of original images of the target domain (vendor) in terms of NIMA scores. Meanwhile, the mr2NST results are also shown to be helpful for the lesion detection in mammograms.

preprint2020arXiv

Peer Offloading in Mobile Edge Computing with Worst-Case Response Time Guarantees

Mobile edge computing (MEC) is a new paradigm that provides cloud computing services at the edge of networks. To achieve better performance with limited computing resources, peer offloading between cooperative edge servers (e.g. MEC- enabled base stations) has been proposed as an effective technique to handle bursty and spatially imbalanced arrival of computation tasks. While various performance metrics of peer offloading policies have been considered in the literatures, the worst-case response time, a common Quality of Service(QoS) requirement in real-time applications, yet receives much less attention. To fill the gap, we formulate the peer offloading problem based on a stochastic arrival model and propose two online algorithms for cases with and without prior knowledge of task arrival rate. Our goal is to maximize the utility function of time-average throughput under constraints of energy consumption and worst-case response time. Both theoretical analysis and numerical results show that our algorithms are able to produce close to optimal performance.

preprint2020arXiv

PointASNL: Robust Point Clouds Processing using Nonlocal Neural Networks with Adaptive Sampling

Raw point clouds data inevitably contains outliers or noise through acquisition from 3D sensors or reconstruction algorithms. In this paper, we present a novel end-to-end network for robust point clouds processing, named PointASNL, which can deal with point clouds with noise effectively. The key component in our approach is the adaptive sampling (AS) module. It first re-weights the neighbors around the initial sampled points from farthest point sampling (FPS), and then adaptively adjusts the sampled points beyond the entire point cloud. Our AS module can not only benefit the feature learning of point clouds, but also ease the biased effect of outliers. To further capture the neighbor and long-range dependencies of the sampled point, we proposed a local-nonlocal (L-NL) module inspired by the nonlocal operation. Such L-NL module enables the learning process insensitive to noise. Extensive experiments verify the robustness and superiority of our approach in point clouds processing tasks regardless of synthesis data, indoor data, and outdoor data with or without noise. Specifically, PointASNL achieves state-of-the-art robust performance for classification and segmentation tasks on all datasets, and significantly outperforms previous methods on real-world outdoor SemanticKITTI dataset with considerate noise. Our code is released through https://github.com/yanx27/PointASNL.

preprint2020arXiv

Re-balancing Variational Autoencoder Loss for Molecule Sequence Generation

Molecule generation is to design new molecules with specific chemical properties and further to optimize the desired chemical properties. Following previous work, we encode molecules into continuous vectors in the latent space and then decode the vectors into molecules under the variational autoencoder (VAE) framework. We investigate the posterior collapse problem of current RNN-based VAEs for molecule sequence generation. For the first time, we find that underestimated reconstruction loss leads to posterior collapse, and provide both theoretical and experimental evidence. We propose an effective and efficient solution to fix the problem and avoid posterior collapse. Without bells and whistles, our method achieves SOTA reconstruction accuracy and competitive validity on the ZINC 250K dataset. When generating 10,000 unique valid SMILES from random prior sampling, it costs JT-VAE1450s while our method only needs 9s. Our implementation is at https://github.com/chaoyan1037/Re-balanced-VAE.

preprint2020arXiv

Task-agnostic Temporally Consistent Facial Video Editing

Recent research has witnessed the advances in facial image editing tasks. For video editing, however, previous methods either simply apply transformations frame by frame or utilize multiple frames in a concatenated or iterative fashion, which leads to noticeable visual flickers. In addition, these methods are confined to dealing with one specific task at a time without any extensibility. In this paper, we propose a task-agnostic temporally consistent facial video editing framework. Based on a 3D reconstruction model, our framework is designed to handle several editing tasks in a more unified and disentangled manner. The core design includes a dynamic training sample selection mechanism and a novel 3D temporal loss constraint that fully exploits both image and video datasets and enforces temporal consistency. Compared with the state-of-the-art facial image editing methods, our framework generates video portraits that are more photo-realistic and temporally smooth.

preprint2020arXiv

Theoretical Analysis of Double Differential Cross Section of Proton, Deuteron and Triton for $p+^7$Li Reaction at 14 MeV

Based on the statistical theory of light nucleus reactions (STLN), the description of the complicated emission processes of proton and light composite charged particles are further improved through considering the effects of Coulomb barriers both in incident and different outgoing reaction channels. And the analysis of the reaction channels including the sequential and simultaneous emission processes for $p + ^7$Li reaction is performed in detail. So the partial spectra of all of outgoing particles are also obtained for different reaction processes. The calculated double differential cross sections of total outgoing proton, deuteron and triton at $E_p = 14$ MeV agree well with the available experimental data for different outgoing angles. The ENDF-6 formatted data, which includes all of the reaction cross sections, elastic angular distributions, double differential cross sections of nucleon and light composite charged particles for $p + ^7$Li reaction, are also obtained by PUNF code.

preprint2016arXiv

Deep Learning At Scale and At Ease

Recently, deep learning techniques have enjoyed success in various multimedia applications, such as image classification and multi-modal data analysis. Large deep learning models are developed for learning rich representations of complex data. There are two challenges to overcome before deep learning can be widely adopted in multimedia and other applications. One is usability, namely the implementation of different models and training algorithms must be done by non-experts without much effort especially when the model is large and complex. The other is scalability, that is the deep learning system must be able to provision for a huge demand of computing resources for training large models with massive datasets. To address these two challenges, in this paper, we design a distributed deep learning platform called SINGA which has an intuitive programming model based on the common layer abstraction of deep learning models. Good scalability is achieved through flexible distributed training architecture and specific optimization techniques. SINGA runs on GPUs as well as on CPUs, and we show that it outperforms many other state-of-the-art deep learning systems. Our experience with developing and training deep learning models for real-life multimedia applications in SINGA shows that the platform is both usable and scalable.

preprint2016arXiv

Feasibility study of online tuning of the luminosity in a circular collider with the robust conjugate direction search method

The robust conjugate direction search (RCDS) method has high tolerance to noise in beam experiments. It has been demonstrated that this method can be used to optimize the machine performance of a light source online. In our study, taking BEPCII as an example, the feasibility of online tuning of the luminosity in a circular collider is explored, through numerical simulation and preliminary online experiments. It is shown that the luminosity that is artificially decreased by a deviation of beam orbital offset from optimal trajectory can be recovered with this method.

preprint2016arXiv

Feature Learning based Deep Supervised Hashing with Pairwise Labels

Recent years have witnessed wide application of hashing for large-scale image retrieval. However, most existing hashing methods are based on hand-crafted features which might not be optimally compatible with the hashing procedure. Recently, deep hashing methods have been proposed to perform simultaneous feature learning and hash-code learning with deep neural networks, which have shown better performance than traditional hashing methods with hand-crafted features. Most of these deep hashing methods are supervised whose supervised information is given with triplet labels. For another common application scenario with pairwise labels, there have not existed methods for simultaneous feature learning and hash-code learning. In this paper, we propose a novel deep hashing method, called deep pairwise-supervised hashing(DPSH), to perform simultaneous feature learning and hash-code learning for applications with pairwise labels. Experiments on real datasets show that our DPSH method can outperform other methods to achieve the state-of-the-art performance in image retrieval applications.

preprint2016arXiv

Service Function Chaining Resource Allocation: A Survey

Service Function Chaining (SFC) is a crucial technology for future Internet. It aims to overcome the limitation of current deployment models which is rigid and static. Application of this technology relies on algorithms that can optimally mapping SFC to substrate network. This category of algorithms is referred as "Service Function Chaining Resource Allocation (SFC-RA)" algorithms or "VNF placement (VNFP)" algorithms. This paper presents a survey of current researches in SFC-RA algorithms. After presenting the formulation and related problems, several variants of SFC-RA problem are summarized. At last, we discussed several future research directions.

preprint2016arXiv

Study on the magnetic measurement results of the injection system for CSNS/RCS

A combination of the H- stripping and phase space painting method is used to accumulate a high intensity beam in the Rapid Cycling Synchrotron (RCS) of the China Spallation Neutron Source (CSNS). The injection system for CSNS/RCS consists of three kinds of magnets: four direct current magnets (BC1-BC4), eight alternating current magnets (BH1-BH4 and BV1-BV4), two septum magnets (ISEP1 and ISEP2). In this paper, the magnetic measurements of the injection system were introduced and the data analysis was processed. The field uniformity and magnetizing curves of these magnets were given, and then the magnetizing fitting equations were obtained.

preprint2016arXiv

The design study for a 500 MeV proton synchrotron with CSNS linac as an injector

Using the China Spallation Neutron Source (CSNS) linac as the injector, a 500 MeV proton synchrotron is proposed for multidisciplinary application, such as biology, material and proton therapy. The synchrotron will deliver proton beam with energy from 80 MeV to 500 MeV. A compact lattice design was worked out, and all the important beam dynamics issues were investigated. The 80 MeV H- beam is stripped and injected into the synchrotron by using multi-turn injection. In order to continuously extraction the proton with small beam loss, the achromatic structure is proposed and slow extraction method with RF knock-out is adopted and optimized.

preprint2016arXiv

The optimization for the conceptual design of a 300 MeV proton synchrotron

A research complex for aerospace radiation effects research has been proposed in Harbin Institute of Technology. Its core part is a proton accelerator complex, which consists of a 10 MeV injector, a 300 MeV synchrotron and beam transport lines. The proton beam extracted from the synchrotron is utilized for the radiation effects research. Based on the conceptual design [1], the design study for optimizing the synchrotron has been done. A new lattice design was worked out, and the multi-turn injection and slow extraction system were optimized with the new lattice design. In order to improve the time structure of the extracted beam, a RF knock-out method is employed. To meet the requirement of accurate control of dose, the frequency of the RF kicker is well investigated.

preprint2015arXiv

AUC-maximized Deep Convolutional Neural Fields for Sequence Labeling

Deep Convolutional Neural Networks (DCNN) has shown excellent performance in a variety of machine learning tasks. This manuscript presents Deep Convolutional Neural Fields (DeepCNF), a combination of DCNN with Conditional Random Field (CRF), for sequence labeling with highly imbalanced label distribution. The widely-used training methods, such as maximum-likelihood and maximum labelwise accuracy, do not work well on highly imbalanced data. To handle this, we present a new training algorithm called maximum-AUC for DeepCNF. That is, we train DeepCNF by directly maximizing the empirical Area Under the ROC Curve (AUC), which is an unbiased measurement for imbalanced data. To fulfill this, we formulate AUC in a pairwise ranking framework, approximate it by a polynomial function and then apply a gradient-based procedure to optimize it. We then test our AUC-maximized DeepCNF on three very different protein sequence labeling tasks: solvent accessibility prediction, 8-state secondary structure prediction, and disorder prediction. Our experimental results confirm that maximum-AUC greatly outperforms the other two training methods on 8-state secondary structure prediction and disorder prediction since their label distributions are highly imbalanced and also have similar performance as the other two training methods on the solvent accessibility prediction problem which has three equally-distributed labels. Furthermore, our experimental results also show that our AUC-trained DeepCNF models greatly outperform existing popular predictors of these three tasks.

preprint2015arXiv

Bermuda: Bidirectional de novo assembly of transcripts with new insights for handling uneven coverage

Motivation: RNA-seq has made feasible the analysis of a whole set of expressed mRNAs. Mapping-based assembly of RNA-seq reads sometimes is infeasible due to lack of high-quality references. However, de novo assembly is very challenging due to uneven expression levels among transcripts and also the read coverage variation within a single transcript. Existing methods either apply de Bruijn graphs of single-sized k-mers to assemble the full set of transcripts, or conduct multiple runs of assembly, but still apply graphs of single-sized k-mers at each run. However, a single k-mer size is not suitable for all the regions of the transcripts with varied coverage. Contribution: This paper presents a de novo assembler Bermuda with new insights for handling uneven coverage. Opposed to existing methods that use a single k-mer size for all the transcripts in each run of assembly, Bermuda self-adaptively uses a few k-mer sizes to assemble different regions of a single transcript according to their local coverage. As such, Bermuda can deal with uneven expression levels and coverage not only among transcripts, but also within a single transcript. Extensive tests show that Bermuda outperforms popular de novo assemblers in reconstructing unevenly-expressed transcripts with longer length, better contiguity and lower redundancy. Further, Bermuda is computationally efficient with moderate memory consumption.

preprint2015arXiv

Electron emission of stripping foil and collimation system for CSNS/RCS

For the Rapid Cycling Synchrotron of the China Spallation Neutron Source (CSNS/RCS), the electron emission plays an important role in the accelerator limitation. The interactions between the proton beam and the stripping foil were studied, and the electron scattering processes were simulated by the ORBIT and FLUKA codes. Then, the electron energy distribution and the electron yielding rate can be given. Furthermore, the interactions between the proton beam and the collimation system were studied, and the electron scattering processes were simulated. Then, the energy distribution of the primary electron emission can be given and the yielding rate of the primary electron can be obtained.

preprint2015arXiv

Geometry-covering Jammer Localization based on Distance Comprehension in Wireless Sensor Networks

Jamming attacks could cause severe damage to Wireless Sensor Networks (WSNs). Once jamming attack occurs, the most urgent work is to get the position information of the jammer. Then safety measures to eliminate the jamming effects can be devised. In this paper, the jammer localization is conducted by geometric covering method to achieve a low energy consumption. And utilizing the power of the jamming signal received by the boundary nodes, a compensating method is composed to reduce the estimating error of the jamming area. At last the localization is conducted by extracting the minimum covering circle of the compensated victim area. Simulations are conducted to test the localization accuracy with the impact of node density, jamming region and radius. Results show that this localization method achieves both good precision and low energy consumption.

preprint2015arXiv

Impedance Measurements of the Extraction Kicker System for the Rapid Cycling Synchrotron of China Spallation Neutron Source

The fast extraction kicker system is one of the most important accelerator components, whose inner structure will be the main source of the impedance in the RCS. It is necessary to understand the kicker impedance before its installation into the tunnel. The conventional and improved wire methods are employed for the benchmarking impedance measurement results. The simulation and measurements confirm that the window-frame ferrite geometry and the end plate are the important structures causing the coupling impedance. The total impedance of the eight modules systems is determined by the scaling law from the measurement and the impedance measurement of the kicker system is summarized.

preprint2015arXiv

Predicting diverse M-best protein contact maps

Protein contacts contain important information for protein structure and functional study, but contact prediction from sequence information remains very challenging. Recently evolutionary coupling (EC) analysis, which predicts contacts by detecting co-evolved residues (or columns) in a multiple sequence alignment (MSA), has made good progress due to better statistical assessment techniques and high-throughput sequencing. Existing EC analysis methods predict only a single contact map for a given protein, which may have low accuracy especially when the protein under prediction does not have a large number of sequence homologs. Analogous to ab initio folding that usually predicts a few possible 3D models for a given protein sequence, this paper presents a novel structure learning method that can predict a set of diverse contact maps for a given protein sequence, in which the best solution usually has much better accuracy than the first one. Our experimental tests show that for many test proteins, the best out of 5 solutions generated by our method has accuracy at least 0.1 better than the first one when the top L/5 or L/10 (L is the sequence length) predicted long-range contacts are evaluated, especially for protein families with a small number of sequence homologs. Our best solutions also have better quality than those generated by the two popular EC methods Evfold and PSICOV.

preprint2015arXiv

Protein Contact Prediction by Integrating Joint Evolutionary Coupling Analysis and Supervised Learning

Protein contacts contain important information for protein structure and functional study, but contact prediction from sequence remains very challenging. Both evolutionary coupling (EC) analysis and supervised machine learning methods are developed to predict contacts, making use of different types of information, respectively. This paper presents a group graphical lasso (GGL) method for contact prediction that integrates joint multi-family EC analysis and supervised learning. Different from existing single-family EC analysis that uses residue co-evolution information in only the target protein family, our joint EC analysis uses residue co-evolution in both the target family and its related families, which may have divergent sequences but similar folds. To implement joint EC analysis, we model a set of related protein families using Gaussian graphical models (GGM) and then co-estimate their precision matrices by maximum-likelihood, subject to the constraint that the precision matrices shall share similar residue co-evolution patterns. To further improve the accuracy of the estimated precision matrices, we employ a supervised learning method to predict contact probability from a variety of evolutionary and non-evolutionary information and then incorporate the predicted probability as prior into our GGL framework. Experiments show that our method can predict contacts much more accurately than existing methods, and that our method performs better on both conserved and family-specific contacts.

preprint2015arXiv

Protein secondary structure prediction using deep convolutional neural fields

Protein secondary structure (SS) prediction is important for studying protein structure and function. When only the sequence (profile) information is used as input feature, currently the best predictors can obtain ~80% Q3 accuracy, which has not been improved in the past decade. Here we present DeepCNF (Deep Convolutional Neural Fields) for protein SS prediction. DeepCNF is a Deep Learning extension of Conditional Neural Fields (CNF), which is an integration of Conditional Random Fields (CRF) and shallow neural networks. DeepCNF can model not only complex sequence-structure relationship by a deep hierarchical architecture, but also interdependency between adjacent SS labels, so it is much more powerful than CNF. Experimental results show that DeepCNF can obtain ~84% Q3 accuracy, ~85% SOV score, and ~72% Q8 accuracy, respectively, on the CASP and CAMEO test proteins, greatly outperforming currently popular predictors. As a general framework, DeepCNF can be used to predict other protein structure properties such as contact number, disorder regions, and solvent accessibility.

preprint2015arXiv

Study on the injection beam commissioning software for CSNS/RCS

The China Spallation Neutron Source (CSNS) accelerator uses H- stripping and phase space painting method of filling large ring acceptance with the linac beam of small emittance. The beam commissioning software system is the key part of CSNS accelerator. The injection beam commissioning software for CSNS contains three parts currently: painting curve control, injection beam control and injection orbit correction. The injection beam control contains two subsections: single bunch beam calculation and LRBT beam control at the foil. The injection orbit correction also contains two subsections: injection orbit correction by the calculation and injection trim power control.

preprint2015arXiv

Study on the transverse painting during the injection process for CSNS/RCS

For the China Spallation Neutron Source (CSNS), a combination of the H- stripping and phase space painting method is used to accumulate a high intensity beam in the Rapid Cycling Synchrotron (RCS). In this paper, firstly, the injection processes with different painting ranges and different painting methods were studied. With the codes ORBIT and MATLAB, the particle distribution and painting image were obtained. Then, the reasonable painting range which is suitable for the aperture size and magnet gap can be selected. Since the real field uniformity of BH3 and BV3 is not completely in conformity with the design requirement, the painting method and painting range also need to be selected to reduce the effects of bad field uniformity.

preprint2015arXiv

Theoretical study of the dual harmonic system and its application on the CSNS/RCS

The dual harmonic system has been widely used in high intensity proton synchrotrons to suppress the space charge effect, as well as reduce the beam loss. To investigate the longitudinal beam dynamics in the dual rf system, the potential well, the sub-buckets in the bunch and the multi-solutions of the phase equation have been studied theoretically. Based on these theoretical studis, the optimization of bunching factor and rf voltage waveform are made for the dual harmonic rf system in the upgrade phase of the CSNS/RCS. In the optimization process, the simulation with space charge effect is done by using a newly developed code C-SCSIM.

preprint2014arXiv

MRFalign: Protein Homology Detection through Alignment of Markov Random Fields

Sequence-based protein homology detection has been extensively studied and so far the most sensitive method is based upon comparison of protein sequence profiles, which are derived from multiple sequence alignment (MSA) of sequence homologs in a protein family. A sequence profile is usually represented as a position-specific scoring matrix (PSSM) or an HMM (Hidden Markov Model) and accordingly PSSM-PSSM or HMM-HMM comparison is used for homolog detection. This paper presents a new homology detection method MRFalign, consisting of three key components: 1) a Markov Random Fields (MRF) representation of a protein family; 2) a scoring function measuring similarity of two MRFs; and 3) an efficient ADMM (Alternating Direction Method of Multipliers) algorithm aligning two MRFs. Compared to HMM that can only model very short-range residue correlation, MRFs can model long-range residue interaction pattern and thus, encode information for the global 3D structure of a protein family. Consequently, MRF-MRF comparison for remote homology detection shall be much more sensitive than HMM-HMM or PSSM-PSSM comparison. Experiments confirm that MRFalign outperforms several popular HMM or PSSM-based methods in terms of both alignment accuracy and remote homology detection and that MRFalign works particularly well for mainly beta proteins. For example, tested on the benchmark SCOP40 (8353 proteins) for homology detection, PSSM-PSSM and HMM-HMM succeed on 48% and 52% of proteins, respectively, at superfamily level, and on 15% and 27% of proteins, respectively, at fold level. In contrast, MRFalign succeeds on 57.3% and 42.5% of proteins at superfamily and fold level, respectively. This study implies that long-range residue interaction patterns are very helpful for sequence-based homology detection. The software is available for download at http://raptorx.uchicago.edu/download/.

preprint2014arXiv

Optimization parameter design for proton irradiation accelerator

The proton irradiation accelerator is widely founded for industry application, and should be designed as compact, reliable, and easy operate. A 10 MeV proton beam is designed to be injected into the slow circulation ring with the repetition rate of 0.5 Hz for accumulation and acceleration, and then the beam with the energy of 300MeV will be slowly extracted by third order resonance method. For getting a higher intensity and more uniform beam, the height of the injection bump is carefully optimised during the injection period. Besides, in order to make the extracted beam with a more uniform distribution, a RF Knock-out method is adopted, and the RF kicker's amplitude is well optimised.

preprint2014arXiv

The study of beam loading effect in the CSNS/RCS

CSNS/RCS accelerates a high-intensity proton beam from 80 MeV to 1.6 GeV. Since the beam current and beam power is high, the beam loading is a severe problem for the stability of the circulating beam in the RCS. To study the beam loading effect in the CSNS/RCS theoretically, the RLC circuit model of the rf cavity, the method of the Fast Fourier Transform and the method of Laplace transform have been employed to obtain the impedance of the rf system, the beam spectrum and the beam-induced voltage, respectively. Based on these physical models, the beam dynamics equations have been revised and a beam loading model has been constructed in the simulation code ORIENT. By using the code, the beam loading effect on the rf system of the CSNS/RCS has been investigated. Some simulation results have been obtained and conclusions have been drawn.

preprint2013arXiv

Analytical Expressions and Numerical simulation of single electron spectroscopy

We use the Monte Carlo method to study the two types of devices used in the technique of single electron spectroscopy and get the C-V curve and I-V curve of them. The results compare well to approximate analytical expressions. Furthermore, with great prospects, we may take into account such effects as cotunneling and coupling between quantum dots through the combination of Monte Carlo method and other numerical methods.

preprint2013arXiv

Optimization Parameter Design of a Circular e+e- Higgs Factory

In this paper we will show a general method of how to make an optimized parameter design of a circular e+e- Higgs Factory by using analytical expression of maximum beam-beam parameter and beamstrahlung beam lifetime started from given design goal and technical limitations. A parameter space has been explored. Based on beam parameters scan and RF parameters scan, a set of optimized parameter designs for 50 km Circular Higgs Factory (CHF) with different RF frequency was proposed.

preprint2013arXiv

The Study on the Effects of Chromaticity and Magnetic Field Tracking Errors at CSNS/RCS

The Rapid Cycling Synchrotron (RCS) is a key component of the China Spallation Neutron Source (CSNS). For this type of high intensity proton synchrotron, the chromaticity, space charge effects and magnetic field tracking errors between the quadrupoles and the dipoles can induce beta function distortion and tune shift, and induce resonances. In this paper the combined effects of chromaticity, magnetic field tracking errors and space charge on beam dynamics at CSNS/RCS are studied systemically. 3-D simulations with different magnetic field tracking errors are performed by using the code ORBIT, and the simulation results are compared with the case without tracking errors.

preprint2012arXiv

Beam losses due to the foil scattering for CSNS/RCS

For the Rapid Cycling Synchrotron of China Spallation Neutron Source (CSNS/RCS), the stripping foil scattering generates the beam halo and gives rise to additional beam losses during the injection process. The interaction between the proton beam and the stripping foil was discussed and the foil scattering was studied. A simple model and the realistic situation of the foil scattering were considered. By using the codes ORBIT and FLUKA, the multi-turn phase space painting injection process with the stripping foil scattering for CSNS/RCS was simulated and the beam losses due to the foil scattering were obtained.

preprint2012arXiv

LogBase: A Scalable Log-structured Database System in the Cloud

Numerous applications such as financial transactions (e.g., stock trading) are write-heavy in nature. The shift from reads to writes in web applications has also been accelerating in recent years. Write-ahead-logging is a common approach for providing recovery capability while improving performance in most storage systems. However, the separation of log and application data incurs write overheads observed in write-heavy environments and hence adversely affects the write throughput and recovery time in the system. In this paper, we introduce LogBase - a scalable log-structured database system that adopts log-only storage for removing the write bottleneck and supporting fast system recovery. LogBase is designed to be dynamically deployed on commodity clusters to take advantage of elastic scaling property of cloud environments. LogBase provides in-memory multiversion indexes for supporting efficient access to data maintained in the log. LogBase also supports transactions that bundle read and write operations spanning across multiple records. We implemented the proposed system and compared it with HBase and a disk-based log-structured record-oriented system modeled after RAMCloud. The experimental results show that LogBase is able to provide sustained write throughput, efficient data access out of the cache, and effective system recovery.

preprint2012arXiv

Study on the injection optimization and transverse coupling for CSNS/RCS

The injection system of the China Spallation Neutron Source uses H- stripping and phase space painting method to fill large ring acceptance with the linac beam of small emittance. The emittance evolution, beam losses, and collimation efficiency during the injection procedures for different injection parameters, such as the injection emittances, starting injection time, twiss parameters and momentum spread, were studied, and then the optimized injection parameters was obtained. In addition, the phase space painting scheme which also affect the emittance evolution and beam losses were simulated and the optimization range of phase space painting were obtained. There will be wobble in the power supply of the injection bumps, and the wobble effects were presented. In order to study the transverse coupling, the injection procedures for different betatron tunes and momentum spreads were studied.

preprint2012arXiv

The effects of injection beam parameters and foil scattering for CSNS/RCS

The China Spallation Neutron Source (CSNS) uses H- stripping and phase space painting method to fill large ring acceptance with the linac beam of small emittance. The dependence of the painting beam on the injection beam parameters was studied for the Rapid Cycling Synchrotron (RCS) of CSNS. The injection processes for different momentum spread, rms emittance of the injection beam, injection beam matching were simulated, then the beam losses, 99% and rms emittances were obtained and the optimized ranges of injection beam parameters were given. The interaction between the H- beam and the stripping foil was studied and the foil scattering was simulated. Then, the stripping efficiency was calculated and the suitable thickness of the stripping foil was obtained. The energy deposition on the foil and the beam losses due to the foil scattering were also studied.

preprint2011arXiv

Cosmological Information in Weak Lensing Peaks

Recent studies have shown that the number counts of convergence peaks N(kappa) in weak lensing (WL) maps, expected from large forthcoming surveys, can be a useful probe of cosmology. We follow up on this finding, and use a suite of WL convergence maps, obtained from ray-tracing N-body simulations, to study (i) the physical origin of WL peaks with different heights, and (ii) whether the peaks contain information beyond the convergence power spectrum P_ell. In agreement with earlier work, we find that high peaks (with amplitudes >~ 3.5 sigma, where sigma is the r.m.s. of the convergence kappa) are typically dominated by a single massive halo. In contrast, medium-height peaks (~0.5-1.5 sigma) cannot be attributed to a single collapsed dark matter halo, and are instead created by the projection of multiple (typically, 4-8) halos along the line of sight, and by random galaxy shape noise. Nevertheless, these peaks dominate the sensitivity to the cosmological parameters w, sigma_8, and Omega_m. We find that the peak height distribution and its dependence on cosmology differ significantly from predictions in a Gaussian random field. We directly compute the marginalized errors on w, sigma_8, and Omega_m from the N(kappa) + P_ell combination, including redshift tomography with source galaxies at z_s=1 and z_s=2. We find that the N(kappa) + P_ell combination has approximately twice the cosmological sensitivity compared to P_ell alone. These results demonstrate that N(kappa) contains non-Gaussian information complementary to the power spectrum.

preprint2011arXiv

Probing Cosmology with Weak Lensing Minkowski Functionals

In this paper, we show that Minkowski Functionals (MFs) of weak gravitational lensing (WL) convergence maps contain significant non-Gaussian, cosmology-dependent information. To do this, we use a large suite of cosmological ray-tracing N-body simulations to create mock WL convergence maps, and study the cosmological information content of MFs derived from these maps. Our suite consists of 80 independent 512^3 N-body runs, covering seven different cosmologies, varying three cosmological parameters Omega_m, w, and sigma_8 one at a time, around a fiducial LambdaCDM model. In each cosmology, we use ray-tracing to create a thousand pseudo-independent 12 deg^2 convergence maps, and use these in a Monte Carlo procedure to estimate the joint confidence contours on the above three parameters. We include redshift tomography at three different source redshifts z_s=1, 1.5, 2, explore five different smoothing scales theta_G=1, 2, 3, 5, 10 arcmin, and explicitly compare and combine the MFs with the WL power spectrum. We find that the MFs capture a substantial amount of information from non-Gaussian features of convergence maps, i.e. beyond the power spectrum. The MFs are particularly well suited to break degeneracies and to constrain the dark energy equation of state parameter w (by a factor of ~ three better than from the power spectrum alone). The non-Gaussian information derives partly from the one-point function of the convergence (through V_0, the "area" MF), and partly through non-linear spatial information (through combining different smoothing scales for V_0, and through V_1 and V_2, the boundary length and genus MFs, respectively). In contrast to the power spectrum, the best constraints from the MFs are obtained only when multiple smoothing scales are combined.

preprint2010arXiv

Hydropathy Conformational Letter and its Substitution Matrix HP-CLESUM: an Application to Protein Structural Alignment

Motivation: Protein sequence world is discrete as 20 amino acids (AA) while its structure world is continuous, though can be discretized into structural alphabets (SA). In order to reveal the relationship between sequence and structure, it is interesting to consider both AA and SA in a joint space. However, such space has too many parameters, so the reduction of AA is necessary to bring down the parameter numbers. Result: We've developed a simple but effective approach called entropic clustering based on selecting the best mutual information between a given reduction of AAs and SAs. The optimized reduction of AA into two groups leads to hydrophobic and hydrophilic. Combined with our SA, namely conformational letter (CL) of 17 alphabets, we get a joint alphabet called hydropathy conformational letter (hp-CL). A joint substitution matrix with (17*2)*(17*2) indices is derived from FSSP. Moreover, we check the three coding systems, say AA, CL and hp-CL against a large database consisting proteins from family to fold, with their performance on the TopK accuracy of both similar fragment pair (SFP) and the neighbor of aligned fragment pair (AFP). The TopK selection is according to the score calculated by the coding system's substitution matrix. Finally, embedding hp-CL in a pairwise alignment algorithm, say CLeFAPS, to replace the original CL, will get an improvement on the HOMSTRAD benchmark.

preprint2009arXiv

LSST Science Book, Version 2.0

A survey that can cover the sky in optical bands over wide fields to faint magnitudes with a fast cadence will enable many of the exciting science opportunities of the next decade. The Large Synoptic Survey Telescope (LSST) will have an effective aperture of 6.7 meters and an imaging camera with field of view of 9.6 deg^2, and will be devoted to a ten-year imaging survey over 20,000 deg^2 south of +15 deg. Each pointing will be imaged 2000 times with fifteen second exposures in six broad bands from 0.35 to 1.1 microns, to a total point-source depth of r~27.5. The LSST Science Book describes the basic parameters of the LSST hardware, software, and observing plans. The book discusses educational and outreach opportunities, then goes on to describe a broad range of science that LSST will revolutionize: mapping the inner and outer Solar System, stellar populations in the Milky Way and nearby galaxies, the structure of the Milky Way disk and halo and other objects in the Local Volume, transient and variable objects both at low and high redshift, and the properties of normal and active galaxies at low and high redshift. It then turns to far-field cosmological topics, exploring properties of supernovae to z~1, strong and weak lensing, the large-scale distribution of galaxies and baryon oscillations, and how these different probes may be combined to constrain cosmological models and the physics of dark energy.

preprint2008arXiv

Challenges to the DGP Model from Horizon-Scale Growth and Geometry

We conduct a Markov Chain Monte Carlo study of the Dvali-Gabadadze-Porrati (DGP) self-accelerating braneworld scenario given the cosmic microwave background (CMB) anisotropy, supernovae and Hubble constant data by implementing an effective dark energy prescription for modified gravity into a standard Einstein-Boltzmann code. We find no way to alleviate the tension between distance measures and horizon scale growth in this model. Growth alterations due to perturbations propagating into the bulk appear as excess CMB anisotropy at the lowest multipoles. In a flat cosmology, the maximum likelihood DGP model is nominally a 5.3 sigma poorer fit than Lambda CDM. Curvature can reduce the tension between distance measures but only at the expense of exacerbating the problem with growth leading to a 4.8 sigma result that is dominated by the low multipole CMB temperature spectrum. While changing the initial conditions to reduce large scale power can flatten the temperature spectrum, this also suppresses the large angle polarization spectrum in violation of recent results from WMAP5. The failure of this model highlights the power of combining growth and distance measures in cosmology as a test of gravity on the largest scales.

preprint2008arXiv

Constraining Cosmology with High Convergence Regions in Weak Lensing Surveys

We propose to use a simple observable, the fractional area of "hot spots" in weak gravitational lensing mass maps which are detected with high significance, to determine background cosmological parameters. Because these high-convergence regions are directly related to the physical nonlinear structures of the universe, they derive cosmological information mainly from the nonlinear regime of density fluctuations. We show that in combination with future cosmic microwave background anisotropy measurements, this method can place constraints on cosmological parameters that are comparable to those from the redshift distribution of galaxy cluster abundances. The main advantage of the statistic proposed in this paper is that projection effects, normally the main source of uncertainty when determining the presence and the mass of a galaxy cluster, here serve as a source of information.

preprint2005arXiv

Weighing Neutrinos with Galaxy Cluster Surveys

Large future galaxy cluster surveys, combined with cosmic microwave background observations, can achieve a high sensitivity to the masses of cosmologically important neutrinos. We show that a weak lensing selected sample of ~100,000 clusters could tighten the current upper bound on the sum of masses of neutrino species by an order of magnitude, to a level of 0.03 eV. Since this statistical sensitivity is below the best existing lower limit on the mass of at least one neutrino species, a future detection is likely, provided that systematic errors can be controlled to a similar level.

Sheng Wang

What is connected

Connect this record

See the researcher in context

Building this map preview

68 published item(s)

AutoResearchClaw: Self-Reinforcing Autonomous Research with Human-AI Collaboration

Hypergraph Pattern Machine: Compositional Tokenization for Higher-Order Interactions

MicroDiffuse3D: A Foundation Model for 3D Microscopy Imaging Restoration

PRISM: Fast Online LLM Serving via Scheduling-Memory Co-design

Relay Buffer Independent Communication over Pooled HBM for Efficient MoE Inference on Ascend

Long time behaviors for the inhomogeneous NLS with a potential in $\mathbb{R}^3$

Adaptive Transfer Learning for Plant Phenotyping

Brain-Aware Replacements for Supervised Contrastive Learning in Detection of Alzheimer's Disease

Estimating risks of option books using neural-SDE market models

Eye-gaze-guided Vision Transformer for Rectifying Shortcut Learning

Follow My Eye: Using Gaze to Supervise Computer-Aided Diagnosis

Graph-in-Graph Network for Automatic Gene Ontology Description Generation

Hedging option books using neural-SDE market models

Joint Progressive and Coarse-to-fine Registration of Brain MRI via Deformation Field Integration and Non-Rigid Feature Fusion

Knee Cartilage Defect Assessment by Graph Representation and Surface Convolution

Non-degeneracy of Stochastic Line Integrals

ProTranslator: zero-shot protein function prediction using textual description

Scattering For three waves Nonlinear Schrödinger System with mass-resonance in 5D

Spin-charge separation in a 1D Fermi gas with tunable interactions

TBI-GAN: An Adversarial Learning Approach for Data Synthesis on Traumatic Brain Segmentation

Towards Accurate Active Camera Localization

Combined Trojan Y Chromosome Strategy and Sterile Insect Technique to Eliminate Mosquitoes: Modelling and Analysis

Field-effect at electrical contacts to two-dimensional materials

Radiation-tolerant high-entropy alloys via interstitial-solute-induced chemical heterogeneities

Analysis of Indexing Structures for Immutable Data

ForkBase: Immutable, Tamper-evident Storage Substrate for Branchable Applications

Label-Driven Reconstruction for Domain Adaptation in Semantic Segmentation

mr2NST: Multi-Resolution and Multi-Reference Neural Style Transfer for Mammography

Peer Offloading in Mobile Edge Computing with Worst-Case Response Time Guarantees

PointASNL: Robust Point Clouds Processing using Nonlocal Neural Networks with Adaptive Sampling

Re-balancing Variational Autoencoder Loss for Molecule Sequence Generation

Task-agnostic Temporally Consistent Facial Video Editing

Theoretical Analysis of Double Differential Cross Section of Proton, Deuteron and Triton for $p+^7$Li Reaction at 14 MeV

Deep Learning At Scale and At Ease

Feasibility study of online tuning of the luminosity in a circular collider with the robust conjugate direction search method

Feature Learning based Deep Supervised Hashing with Pairwise Labels

Service Function Chaining Resource Allocation: A Survey

Study on the magnetic measurement results of the injection system for CSNS/RCS

The design study for a 500 MeV proton synchrotron with CSNS linac as an injector

The optimization for the conceptual design of a 300 MeV proton synchrotron

AUC-maximized Deep Convolutional Neural Fields for Sequence Labeling

Bermuda: Bidirectional de novo assembly of transcripts with new insights for handling uneven coverage

Electron emission of stripping foil and collimation system for CSNS/RCS

Geometry-covering Jammer Localization based on Distance Comprehension in Wireless Sensor Networks

Impedance Measurements of the Extraction Kicker System for the Rapid Cycling Synchrotron of China Spallation Neutron Source

Predicting diverse M-best protein contact maps

Protein Contact Prediction by Integrating Joint Evolutionary Coupling Analysis and Supervised Learning

Protein secondary structure prediction using deep convolutional neural fields

Study on the injection beam commissioning software for CSNS/RCS

Study on the transverse painting during the injection process for CSNS/RCS

Theoretical study of the dual harmonic system and its application on the CSNS/RCS

MRFalign: Protein Homology Detection through Alignment of Markov Random Fields

Optimization parameter design for proton irradiation accelerator

The study of beam loading effect in the CSNS/RCS

Analytical Expressions and Numerical simulation of single electron spectroscopy

Optimization Parameter Design of a Circular e+e- Higgs Factory

The Study on the Effects of Chromaticity and Magnetic Field Tracking Errors at CSNS/RCS

Beam losses due to the foil scattering for CSNS/RCS

LogBase: A Scalable Log-structured Database System in the Cloud

Study on the injection optimization and transverse coupling for CSNS/RCS

The effects of injection beam parameters and foil scattering for CSNS/RCS

Cosmological Information in Weak Lensing Peaks

Probing Cosmology with Weak Lensing Minkowski Functionals

Hydropathy Conformational Letter and its Substitution Matrix HP-CLESUM: an Application to Protein Structural Alignment

LSST Science Book, Version 2.0

Challenges to the DGP Model from Horizon-Scale Growth and Geometry

Constraining Cosmology with High Convergence Regions in Weak Lensing Surveys

Weighing Neutrinos with Galaxy Cluster Surveys