BZPEERPapers, people, evidence and research paths
Menu

Discover

SearchFind papers, people, topics, institutions and opportunities from one place.GraphSee how papers, authors, topics and evidence connect.

Research tools

HomeContinue from the papers, people and topics that matter now.CollectionsTurn saved papers into reading lists, evidence maps and shareable context.ResearchReconstruct the path that led you from search to paper to question.CollaborationMove from a paper, topic or expert request into a focused thread.PublishCreate full-text research articles with metadata, formulas and versions.

Network

ExpertsFind specialists whose work and reviews explain why they can help.CommunitiesJoin research groups organized around papers, topics and shared evidence.

Opportunities

ListingsBrowse roles, calls and collaborations connected to real research context.

Account

Log inReturn to your saved papers, graph views and active threads.Create accountStart saving papers, building a research trail and joining teams.
Log inCreate account

Source author record

Jie Feng

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

ResearcherUnclaimed source record
Computer VisionMachine Learningphysics.flu-dynphysics.opticsArtificial Intelligenceastro-ph.HEcond-mat.mtrl-scihep-phphysics.plasm-phSocial and Information Networksastro-ph.IMcond-mat.softeess.IVhep-exnucl-exphysics.ao-ph

Catalog footprint

What is connected

19works
16topics
4close collaborators

Actions

Connect this record

Log in to claim
Open graphBrowse works

Research graph

See the researcher in context

Open full explorer

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.

Published work

19 published item(s)

preprint2026arXiv

LMM-Track4D: Eliciting 4D Dynamic Reasoning in LMMs via Trajectory-Grounded Dialogue

Recent large multimodal models (LMMs) have become increasingly capable on image and video understanding, yet still struggle to sustain 4D continuous spatiotemporal dynamic reasoning. To study this capability gap, we formulate trajectory-grounded multi-turn spatiotemporal dialogue, a new task in which a model must answer spatiotemporal queries while returning structured 3D target trajectories over an entire short clip or a specified segment of a longer clip, and introduce Track4D-Bench, a benchmark with 526 clip-level dialogue samples spanning 23.5k frames and 7.5k object annotations, for training and evaluation. Building on this task, we propose LMM-Track4D, which combines RTGE (Ray--Time Geometry Encoding), a dedicated streaming state token TRK for long-horizon dynamic propagation, and an Object-Slot Kinematic, Residual-Anchor (OSK-RA) decoder for stable 4-step 3D state estimation under occlusion and viewpoint variation. Experiments on Track4D-Bench show consistent improvements over strong baselines, suggesting that explicit dynamic state modeling is a useful design principle for eliciting 4D dynamic reasoning in LMMs. Our code and dataset will be publicly available at https://github.com/mikubaka88/LMM-Track4D.

0 citations0 reviewsNo code linkedOpen access
preprint2026arXiv

Tabletop X-ray ghost video of moving objects

X-ray imaging is widely employed in clinical medicine, industrial inspection, and various scientific research fields. Unfortunately, most currently used X-ray two-dimensional (2D) detectors suffer from a fundamental trade-off between the number of pixels and readout time, making them unsuitable for fast moving objects imaging, as well as the readout dead time causes frame losses. X-ray ghost imaging (XGI) offers an alternative approach to image an object using only a highly sensitive single-pixel detector. However, a critical limitation of existing XGI methods is the excessive total acquisition time required, rendering it impractical for real applications. In this paper, we propose a rapid spatial modulation scheme based on random binary patterns encoded onto a fast-spinning mask. Clear X-ray visualization of moving objects is demonstrated with imaging rates up to 200 frames per second with a resolution of 225 um. For the first time, our method has greatly improved the XGI imaging speed and paves the way for X-ray imaging application of motion objects, such as the inspection of rotating aero-engines and in vivo medical imaging.

0 citations0 reviewsNo code linkedOpen access
preprint2025arXiv

Oscillatory flows in three-dimensional deformable microchannels

Deformable microchannels emulate a key characteristic of soft biological systems and flexible engineering devices: the flow-induced deformation of the conduit due to slow viscous flow within. Elucidating the two-way coupling between oscillatory flow and deformation of a three-dimensional (3D) rectangular channel is crucial for designing lab- and organ-on-a-chip microsystems and eventually understanding flow-structure instabilities that can enhance mixing and transport. To this end, we determine the axial variations of the primary flow, pressure, and deformation for Newtonian fluids in the canonical geometry of a slender (long) and shallow (wide) 3D rectangular channel with a deformable top wall under the assumption of weak compliance and without restriction on the oscillation frequency (\textit{i.e.}, on the Womersley number). Unlike rigid conduits, the pressure distribution is not linear with the axial coordinate. To validate this prediction, we design a PDMS-based experimental platform with a speaker-based flow-generation apparatus and a pressure acquisition system with multiple ports along the axial length of the channel. The experimental measurements show good agreement with the predicted pressure profiles across a wide range of the key dimensionless quantities: the Womersley number, the compliance number, and the elastoviscous number. Finally, we explore how the nonlinear flow-deformation coupling leads to self-induced streaming (rectification of the oscillatory flow). Following Zhang and Rallabandi (\textit{J.\ Fluid Mech.}, vol.~996, 2024, A16), we develop a theory for the cycle-averaged pressure based on the primary problem's solution, and we validate the predictions for the axial distribution of the streaming pressure against the experimental measurements.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

CoSimGNN: Towards Large-scale Graph Similarity Computation

The ability to compute similarity scores between graphs based on metrics such as Graph Edit Distance (GED) is important in many real-world applications. Computing exact GED values is typically an NP-hard problem and traditional algorithms usually achieve an unsatisfactory trade-off between accuracy and efficiency. Recently, Graph Neural Networks (GNNs) provide a data-driven solution for this task, which is more efficient while maintaining prediction accuracy in small graph (around 10 nodes per graph) similarity computation. Existing GNN-based methods, which either respectively embeds two graphs (lack of low-level cross-graph interactions) or deploy cross-graph interactions for whole graph pairs (redundant and time-consuming), are still not able to achieve competitive results when the number of nodes in graphs increases. In this paper, we focus on similarity computation for large-scale graphs and propose the "embedding-coarsening-matching" framework CoSimGNN, which first embeds and coarsens large graphs with adaptive pooling operation and then deploys fine-grained interactions on the coarsened graphs for final similarity scores. Furthermore, we create several synthetic datasets which provide new benchmarks for graph similarity computation. Detailed experiments on both synthetic and real-world datasets have been conducted and CoSimGNN achieves the best performance while the inference time is at most 1/3 of that of previous state-of-the-art.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Femtosecond pumping of nuclear isomeric states by the Coulomb collision of ions with quivering electrons

Efficient production of nuclear isomers is critical for pioneering applications, like nuclear clocks, nuclear batteries, clean nuclear energy, and nuclear γ-ray lasers. However, due to small production cross sections and quick decays, it is extremely difficult to acquire a significant amount of isomers with short lifetimes via traditional accelerators or reactors because of low beam intensity. Here, for the first time, we experimentally present femtosecond pumping of nuclear isomeric states by the Coulomb excitation of ions with the quivering electrons induced by laser fields. Nuclei populated on the third excited state of 83Kr are generated with a peak efficiency of 2.34*10^15 particles=s from a tabletop hundred-TW laser system. It can be explained by the Coulomb excitation of ions with the quivering electrons during the interaction between laser pulses and clusters at nearly solid densities. This efficient and universal production method can be widely used for pumping isotopes with excited state lifetimes down to picoseconds, and could be a benefit for fields like nuclear transition mechanisms and nuclear γ-ray lasers.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Laser plasma accelerated ultra-intense electron beam for efficiently exciting nuclear isomers

Utilizing laser plasma wakefield to accelerate ultra-high charge electron beam is critical for many pioneering applications, for example to efficiently produce nuclear isomers with short lifetimes which may be widely used. However, because of the beam loading effect, electron charge in a single plasma bubble is limited in level of hundreds picocoulomb. Here, we experimentally present that a hundred kilo-ampere, twenty nanocoulomb, tens of MeV collimated electron beam is produced from a chain of wakefield acceleration, via a tightly focused intense laser pulse transversely matched in dense plasma. This ultra-intense electron beam ascribes to a novel efficient injection that the nitrogen atom inner shell electrons are ionized and continuously injected into multiple plasma bubbles. This intense electron beam has been utilized to exciting nuclear isomers with an ultra-high peak efficiency of $1.76\times10^{15}$ particles/s via photonuclear reactions. This efficient production method of isomers can be widely used for pumping isotopes with excited state lifetimes down to picosecond, which is benefit for deep understanding nuclear transition mechanisms and stimulating gamma-ray lasers.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

SparseDet: Towards End-to-End 3D Object Detection

In this paper, we propose SparseDet for end-to-end 3D object detection from point cloud. Existing works on 3D object detection rely on dense object candidates over all locations in a 3D or 2D grid following the mainstream methods for object detection in 2D images. However, this dense paradigm requires expertise in data to fulfill the gap between label and detection. As a new detection paradigm, SparseDet maintains a fixed set of learnable proposals to represent latent candidates and directly perform classification and localization for 3D objects through stacked transformers. It demonstrates that effective 3D object detection can be achieved with none of post-processing such as redundant removal and non-maximum suppression. With a properly designed network, SparseDet achieves highly competitive detection accuracy while running with a more efficient speed of 34.5 FPS. We believe this end-to-end paradigm of SparseDet will inspire new thinking on the sparsity of 3D object detection.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Water-to-air transfer of nano/micro-sized particulates: enrichment effect in bubble bursting jet drops

Bubbles dispersed in liquids are widely present in many natural and industrial processes, and play a key role in mediating mass transfer during their lifetime from formation to rising to bursting. In particular, nano/micro-sized particulates and organisms, present in the bulk water can be highly enriched in the jet drops ejected during bubble bursting, impacting global climate and public health. However, the detailed mechanism of this enrichment remains obscure, with the enrichment factor being difficult to predict. Here, we experimentally investigate the enrichment of nano/micro-sized particles in bubble bursting jet drops and highlight the underlying hydrodynamic mechanism, combining the effects of bubble scavenge and bursting on the transport of particles. Scaling laws for the enrichment factor are subsequently proposed that describe both our and prior experimental results reasonably well. Our study may provide new insights for water-to-air transfer of microbes related to bubble bursting.

0 citations0 reviewsNo code linkedOpen access
preprint2021arXiv

AttnMove: History Enhanced Trajectory Recovery via Attentional Network

A considerable amount of mobility data has been accumulated due to the proliferation of location-based service. Nevertheless, compared with mobility data from transportation systems like the GPS module in taxis, this kind of data is commonly sparse in terms of individual trajectories in the sense that users do not access mobile services and contribute their data all the time. Consequently, the sparsity inevitably weakens the practical value of the data even it has a high user penetration rate. To solve this problem, we propose a novel attentional neural network-based model, named AttnMove, to densify individual trajectories by recovering unobserved locations at a fine-grained spatial-temporal resolution. To tackle the challenges posed by sparsity, we design various intra- and inter- trajectory attention mechanisms to better model the mobility regularity of users and fully exploit the periodical pattern from long-term history. We evaluate our model on two real-world datasets, and extensive results demonstrate the performance gain compared with the state-of-the-art methods. This also shows that, by providing high-quality mobility data, our model can benefit a variety of mobility-oriented down-stream applications.

0 citations0 reviewsNo code linkedOpen access
preprint2021arXiv

Graph Partitioning and Graph Neural Network based Hierarchical Graph Matching for Graph Similarity Computation

Graph similarity computation aims to predict a similarity score between one pair of graphs to facilitate downstream applications, such as finding the most similar chemical compounds similar to a query compound or Fewshot 3D Action Recognition. Recently, some graph similarity computation models based on neural networks have been proposed, which are either based on graph-level interaction or node-level comparison. However, when the number of nodes in the graph increases, it will inevitably bring about reduced representation ability or high computation cost. Motivated by this observation, we propose a graph partitioning and graph neural network-based model, called PSimGNN, to effectively resolve this issue. Specifically, each of the input graphs is partitioned into a set of subgraphs to extract the local structural features directly. Next, a novel graph neural network with an attention mechanism is designed to map each subgraph into an embedding vector. Some of these subgraph pairs are automatically selected for node-level comparison to supplement the subgraph-level embedding with fine-grained information. Finally, coarse-grained interaction information among subgraphs and fine-grained comparison information among nodes in different subgraphs are integrated to predict the final similarity score. Experimental results on graph datasets with different graph sizes demonstrate that PSimGNN outperforms state-of-the-art methods in graph similarity computation tasks using approximate Graph Edit Distance (GED) as the graph similarity metric.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

End-to-end Optimized Video Compression with MV-Residual Prediction

We present an end-to-end trainable framework for P-frame compression in this paper. A joint motion vector (MV) and residual prediction network MV-Residual is designed to extract the ensembled features of motion representations and residual information by treating the two successive frames as inputs. The prior probability of the latent representations is modeled by a hyperprior autoencoder and trained jointly with the MV-Residual network. Specially, the spatially-displaced convolution is applied for video frame prediction, in which a motion kernel for each pixel is learned to generate predicted pixel by applying the kernel at a displaced location in the source image. Finally, novel rate allocation and post-processing strategies are used to produce the final compressed bits, considering the bits constraint of the challenge. The experimental results on validation set show that the proposed optimized framework can generate the highest MS-SSIM for P-frame compression competition.

0 citations0 reviewsNo code linkedOpen access
preprint2019arXiv

A Genetic Algorithm for Astroparticle Physics Studies

Precision measurements of charged cosmic rays have recently been carried out by space-born (e.g. AMS-02), or ground experiments (e.g. HESS). These measured data are important for the studies of astro-physical phenomena, including supernova remnants, cosmic ray propagation, solar physics and dark matter. Those scenarios usually contain a number of free parameters that need to be adjusted by observed data. Some techniques, such as Markov Chain Monte Carlo and MultiNest, are developed in order to solve the above problem. However, it is usually required a computing farm to apply those tools. In this paper, a genetic algorithm for finding the optimum parameters for cosmic ray injection and propagation is presented. We find that this algorithm gives us the same best fit results as the Markov Chain Monte Carlo but consuming less computing power by nearly 2 orders of magnitudes.

0 citations0 reviewsNo code linkedOpen access
preprint2016arXiv

Deep Image Set Hashing

In applications involving matching of image sets, the information from multiple images must be effectively exploited to represent each set. State-of-the-art methods use probabilistic distribution or subspace to model a set and use specific distance measure to compare two sets. These methods are slow to compute and not compact to use in a large scale scenario. Learning-based hashing is often used in large scale image retrieval as they provide a compact representation of each sample and the Hamming distance can be used to efficiently compare two samples. However, most hashing methods encode each image separately and discard knowledge that multiple images in the same set represent the same object or person. We investigate the set hashing problem by combining both set representation and hashing in a single deep neural network. An image set is first passed to a CNN module to extract image features, then these features are aggregated using two types of set feature to capture both set specific and database-wide distribution information. The computed set feature is then fed into a multilayer perceptron to learn a compact binary embedding. Triplet loss is used to train the network by forming set similarity relations using class labels. We extensively evaluate our approach on datasets used for image matching and show highly competitive performance compared to state-of-the-art methods.

0 citations0 reviewsNo code linkedOpen access
preprint2016arXiv

Pulsar interpretation of the lepton spectra measured by AMS-02

AMS-02 recently published its lepton spectra measurement. The results show that the positron fraction no longer increases above $\sim$200 GeV. The aim of this work is to investigate the possibility that the excess of positron fraction is due to pulsars. Nearby known pulsars from ATNF catalogue are considered as a possible primary positron source of the high energy positrons. We find that the pulsars with age $T\simeq (0.45\sim4.5)\times10^{5}$ yr and distance $d<0.5$ kpc can explain the behavior of positron fraction of AMS-02 in the range of high energy. We show that each of the four pulsars --- Geminga, J1741-2054, Monogem and J0942-5552 --- is able to be a single source satisfying all considered physical requirements. We also discuss the possibility that these high energy $e^{\pm}$ are from multiple pulsars. The multiple pulsars contribution predicts a positron fraction with some structures at higher energies.

0 citations0 reviewsNo code linkedOpen access
preprint2015arXiv

Deposition of quantum dots in a capillary tube

The ability to assemble nanomaterials, such as quantum dots, enables the creation of functional devices that present unique optical and electronic properties. For instance, light-emitting diodes with exceptional color purity can be printed via the evaporative-driven assembly of quantum dots. Nevertheless, current studies of the colloidal deposition of quantum dots have been limited to the surfaces of a planar substrate. Here, we investigate the evaporation-driven assembly of quantum dots inside a confined cylindrical geometry. Specifically, we observe distinct deposition patterns, such as banding structures along the length of a capillary tube. Such coating behavior can be influenced by the evaporation speed as well as the concentration of quantum dots. Understanding the factors governing the coating process can provide a means to control the assembly of quantum dots inside a capillary tube, ultimately enabling the creation of novel photonic devices.

0 citations0 reviewsNo code linkedOpen access
preprint2014arXiv

Learning to Rank Binary Codes

Binary codes have been widely used in vision problems as a compact feature representation to achieve both space and time advantages. Various methods have been proposed to learn data-dependent hash functions which map a feature vector to a binary code. However, considerable data information is inevitably lost during the binarization step which also causes ambiguity in measuring sample similarity using Hamming distance. Besides, the learned hash functions cannot be changed after training, which makes them incapable of adapting to new data outside the training data set. To address both issues, in this paper we propose a flexible bitwise weight learning framework based on the binary codes obtained by state-of-the-art hashing methods, and incorporate the learned weights into the weighted Hamming distance computation. We then formulate the proposed framework as a ranking problem and leverage the Ranking SVM model to offline tackle the weight learning. The framework is further extended to an online mode which updates the weights at each time new data comes, thereby making it scalable to large and dynamic data sets. Extensive experimental results demonstrate significant performance gains of using binary codes with bitwise weighting in image retrieval tasks. It is appealing that the online weight learning leads to comparable accuracy with its offline counterpart, which thus makes our approach practical for realistic applications.

0 citations0 reviewsNo code linkedOpen access
preprint2013arXiv

Nanoemulsions obtained via bubble bursting at a compound interface

The bursting of bubbles at an air/liquid interface is a familiar occurrence important to foam stability, cell cultures in bioreactors and mass transfer between the sea and atmosphere. Here we document the hitherto unreported formation and dispersal into the water column of submicrometre oil droplets following bubble bursting at a compound air/oil/water-with-surfactant interface. We show that dispersal results from the detachment of an oil spray from the bottom of the bubble towards water during bubble collapse. We provide evidence that droplet size is selected by physicochemical interactions between oil molecules and the surfactants rather than by hydrodynamic effects. We illustrate the unrecognized role that this dispersal mechanism may play in the fate of the sea surface micro-layer and of pollutant spills by dispersing petroleum in the water column. Finally, our system provides an energy-efficient route, with potential upscalability and wide applicability, for applications in drug delivery, food production and material science, which we demonstrate by producing polymeric nanoparticles.

0 citations0 reviewsNo code linkedOpen access
preprint2011arXiv

A Non-linearized PLS Model Based on Multivariate Dominant Factor for Laser-induced Breakdown Spectroscopy Measurements

A multivariate dominant factor based non-linearized PLS model is proposed. The intensities of different lines were taken to construct a multivariate dominant factor model, which describes the dominant concentration information of the measured species. In constructing such a multivariate model, non-linear transformation of multi characteristic line intensities according to the physical mechanisms of lased induced plasma spectrum were made, combined with linear-correlation-based PLS method, to model the nonlinear self-absorption and inter-element interference effects. This enables the linear PLS method to describe non-linear relationship more accurately and provides the statistics-based PLS method with physical backgrounds. Moreover, a secondary PLS is applied utilizing the whole spectra information to further correct the model results. Experiments were conducted using standard brass samples. Taylor expansion was applied to make the nonlinear transformation to describe the self-absorption effect of Cu. Then, line intensities of another two elements, Pb and Zn, were taken into account for inter-element interference. The proposed method shows a significant improvement when compared with conventional PLS model. Results also show that, even compared with the already-improved baseline dominant-factor-based PLS model, the present PLS model based on the multivariate dominant factor yields the same calibration quality (R2=0.999) while decreasing the RMSEP from 2.33% to 1.97%. The overall RMSE was also improved to 1.05% from 1.27%.

0 citations0 reviewsNo code linkedOpen access
preprint2010arXiv

A Novel Multivariate Model Based on Dominant Factor for Laser-induced Breakdown Spectroscopy Measurements

This paper presents a new approach of applying partial least squares method combined with a physical principle based dominant factor. The characteristic line intensity of the specific element was taken to build up the dominant factor to reflect the major elemental concentration and partial least squares (PLS) approach was then applied to further improve the model accuracy. The deviation evolution of characteristic line intensity from the ideal condition was depicted and according to the deviation understanding, efforts were taken to model the non-linear self-absorption and inter-element interference effects to improve the accuracy of dominant factor model. With a dominant factor to carry the main quantitative information, the novel multivariate model combines advantages of both the conventional univariate and PLS models and partially avoids the overuse of the unrelated noise in the spectrum for PLS application. The dominant factor makes the combination model more robust over a wide concentration range and PLS application improves the model accuracy for samples with matrices within the calibration sample set. Results show that RMSEP of the final dominant factor based PLS model decreased to 2.33% from 5.25% when using the conventional PLS approach with full spectral information. Furthermore, with the development in understanding the physics of the laser-induced plasma, there is potential to easily improve the accuracy of the dominant factor model as well as the proposed novel multivariate model.

0 citations0 reviewsNo code linkedOpen access