Searcharxiv⌕ Search

arXiv subjects

Jun Zhang

Publications and source records attributed to Jun Zhang.

At least 631 records · Page 35Linked to original sources
arXiv2023

Forensic Histopathological Recognition via a Context-Aware MIL Network Powered by Self-Supervised Contrastive Learning

Forensic pathology is critical in analyzing death manner and time from the microscopic aspect to assist in the establishment of reliable factual bases for criminal investigation. In practice, even the manual differentiation between different postmortem organ tissues is challenging and relies on expertise, considering that changes like putrefaction and autolysis could significantly change typical histopathological appearance. Developing AI-based computational pathology techniques to assist forensic pathologists is practically meaningful, which requires reliable discriminative representation learning to capture tissues' fine-grained postmortem patterns. To this end, we propose a framework called FPath, in which a dedicated self-supervised contrastive learning strategy and a context-aware multiple-instance learning (MIL) block are designed to learn discriminative representations from postmortem histopathological images acquired at varying magnification scales. Our self-supervised learning step leverages multiple complementary contrastive losses and regularization terms to train a double-tier backbone for fine-grained and informative patch/instance embedding. Thereafter, the context-aware MIL adaptively distills from the local instances a holistic bag/image-level representation for the recognition task. On a large-scale database of $19,607$ experimental rat postmortem images and $3,378$ real-world human decedent images, our FPath led to state-of-the-art accuracy and promising cross-domain generalization in recognizing seven different postmortem tissues. The source code will be released on \href{https://github.com/ladderlab-xjtu/forensic_pathology}{https://github.com/ladderlab-xjtu/forensic\_pathology}.

cs.CV↗
arXiv2023

Dynamic Low-Rank Instance Adaptation for Universal Neural Image Compression

The latest advancements in neural image compression show great potential in surpassing the rate-distortion performance of conventional standard codecs. Nevertheless, there exists an indelible domain gap between the datasets utilized for training (i.e., natural images) and those utilized for inference (e.g., artistic images). Our proposal involves a low-rank adaptation approach aimed at addressing the rate-distortion drop observed in out-of-domain datasets. Specifically, we perform low-rank matrix decomposition to update certain adaptation parameters of the client's decoder. These updated parameters, along with image latents, are encoded into a bitstream and transmitted to the decoder in practical scenarios. Due to the low-rank constraint imposed on the adaptation parameters, the resulting bit rate overhead is small. Furthermore, the bit rate allocation of low-rank adaptation is \emph{non-trivial}, considering the diverse inputs require varying adaptation bitstreams. We thus introduce a dynamic gating network on top of the low-rank adaptation method, in order to decide which decoder layer should employ adaptation. The dynamic adaptation network is optimized end-to-end using rate-distortion loss. Our proposed method exhibits universality across diverse image datasets. Extensive results demonstrate that this paradigm significantly mitigates the domain gap, surpassing non-adaptive methods with an average BD-rate improvement of approximately $19\%$ across out-of-domain images. Furthermore, it outperforms the most advanced instance adaptive methods by roughly $5\%$ BD-rate. Ablation studies confirm our method's ability to universally enhance various image compression architectures.

eess.IV↗
arXiv2023

IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image Diffusion Models

Recent years have witnessed the strong power of large text-to-image diffusion models for the impressive generative capability to create high-fidelity images. However, it is very tricky to generate desired images using only text prompt as it often involves complex prompt engineering. An alternative to text prompt is image prompt, as the saying goes: "an image is worth a thousand words". Although existing methods of direct fine-tuning from pretrained models are effective, they require large computing resources and are not compatible with other base models, text prompt, and structural controls. In this paper, we present IP-Adapter, an effective and lightweight adapter to achieve image prompt capability for the pretrained text-to-image diffusion models. The key design of our IP-Adapter is decoupled cross-attention mechanism that separates cross-attention layers for text features and image features. Despite the simplicity of our method, an IP-Adapter with only 22M parameters can achieve comparable or even better performance to a fully fine-tuned image prompt model. As we freeze the pretrained diffusion model, the proposed IP-Adapter can be generalized not only to other custom models fine-tuned from the same base model, but also to controllable generation using existing controllable tools. With the benefit of the decoupled cross-attention strategy, the image prompt can also work well with the text prompt to achieve multimodal image generation. The project page is available at \url{https://ip-adapter.github.io}.

cs.CV↗
arXiv2023

Feature Matching Data Synthesis for Non-IID Federated Learning

Federated learning (FL) has emerged as a privacy-preserving paradigm that trains neural networks on edge devices without collecting data at a central server. However, FL encounters an inherent challenge in dealing with non-independent and identically distributed (non-IID) data among devices. To address this challenge, this paper proposes a hard feature matching data synthesis (HFMDS) method to share auxiliary data besides local models. Specifically, synthetic data are generated by learning the essential class-relevant features of real samples and discarding the redundant features, which helps to effectively tackle the non-IID issue. For better privacy preservation, we propose a hard feature augmentation method to transfer real features towards the decision boundary, with which the synthetic data not only improve the model generalization but also erase the information of real features. By integrating the proposed HFMDS method with FL, we present a novel FL framework with data augmentation to relieve data heterogeneity. The theoretical analysis highlights the effectiveness of our proposed data synthesis method in solving the non-IID challenge. Simulation results further demonstrate that our proposed HFMDS-FL algorithm outperforms the baselines in terms of accuracy, privacy preservation, and computational cost on various benchmark datasets.

cs.LG↗
arXiv2023

model-based script synthesis for fuzzing

Kernel fuzzing is important for finding critical kernel vulnerabilities. Close-source (e.g., Windows) operating system kernel fuzzing is even more challenging due to the lack of source code. Existing approaches fuzz the kernel by modeling syscall sequences from traces or static analysis of system codes. However, a common limitation is that they do not learn and mutate the syscall sequences to reach different kernel states, which can potentially result in more bugs or crashes. In this paper, we propose WinkFuzz, an approach to learn and mutate traced syscall sequences in order to reach different kernel states. WinkFuzz learns syscall dependencies from the trace, identifies potential syscalls in the trace that can have dependent subsequent syscalls, and applies the dependencies to insert more syscalls while preserving the dependencies into the trace. Then WinkFuzz fuzzes the synthesized new syscall sequence to find system crashes. We applied WinkFuzz to four seed applications and found a total increase in syscall number of 70.8\%, with a success rate of 61\%, within three insert levels. The average time for tracing, dependency analysis, recovering model script, and synthesizing script was 600, 39, 34, and 129 seconds respectively. The instant fuzzing rate is 3742 syscall executions per second. However, the average fuzz efficiency dropped to 155 syscall executions per second when the initializing time, waiting time, and other factors were taken into account. We fuzzed each seed application for 24 seconds and, on average, obtained 12.25 crashes within that time frame.

cs.CR↗
arXiv2023

Binary Federated Learning with Client-Level Differential Privacy

Federated learning (FL) is a privacy-preserving collaborative learning framework, and differential privacy can be applied to further enhance its privacy protection. Existing FL systems typically adopt Federated Average (FedAvg) as the training algorithm and implement differential privacy with a Gaussian mechanism. However, the inherent privacy-utility trade-off in these systems severely degrades the training performance if a tight privacy budget is enforced. Besides, the Gaussian mechanism requires model weights to be of high-precision. To improve communication efficiency and achieve a better privacy-utility trade-off, we propose a communication-efficient FL training algorithm with differential privacy guarantee. Specifically, we propose to adopt binary neural networks (BNNs) and introduce discrete noise in the FL setting. Binary model parameters are uploaded for higher communication efficiency and discrete noise is added to achieve the client-level differential privacy protection. The achieved performance guarantee is rigorously proved, and it is shown to depend on the level of discrete noise. Experimental results based on MNIST and Fashion-MNIST datasets will demonstrate that the proposed training algorithm achieves client-level privacy protection with performance gain while enjoying the benefits of low communication overhead from binary model updates.

cs.LG↗
arXiv2023

SemDiff: Binary Similarity Detection by Diffing Key-Semantics Graphs

Binary similarity detection is a critical technique that has been applied in many real-world scenarios where source code is not available, e.g., bug search, malware analysis, and code plagiarism detection. Existing works are ineffective in detecting similar binaries in cases where different compiling optimizations, compilers, source code versions, or obfuscation are deployed. We observe that all the cases do not change a binary's key code behaviors although they significantly modify its syntax and structure. With this key observation, we extract a set of key instructions from a binary to capture its key code behaviors. By detecting the similarity between two binaries' key instructions, we can address well the ineffectiveness limitation of existing works. Specifically, we translate each extracted key instruction into a self-defined key expression, generating a key-semantics graph based on the binary's control flow. Each node in the key-semantics graph denotes a key instruction, and the node attribute is the key expression. To quantify the similarity between two given key-semantics graphs, we first serialize each graph into a sequence of key expressions by topological sort. Then, we tokenize and concatenate key expressions to generate token lists. We calculate the locality-sensitive hash value for all token lists and quantify their similarity. %We implement a prototype, called SemDiff, consisting of two modules: graph generation and graph diffing. The first module generates a pair of key-semantics graphs and the second module diffs the graphs. Our evaluation results show that overall, SemDiff outperforms state-of-the-art tools when detecting the similarity of binaries generated from different optimization levels, compilers, and obfuscations. SemDiff is also effective for library version search and finding similar vulnerabilities in firmware.

cs.CR↗
arXiv2023

$\mathrm{H}α$ chromospheric activity of F-, G-, and K-type stars observed by the LAMOST Medium-Resolution Spectroscopic Survey

The distribution of stellar $\mathrm{H}α$ chromospheric activity with respect to stellar atmospheric parameters (effective temperature $T_\mathrm{eff}$, surface gravity $\log\,g$, and metallicity $\mathrm{[Fe/H]}$) and main-sequence/giant categories is investigated for the F-, G-, and K-type stars observed by the LAMOST Medium-Resolution Spectroscopic Survey (MRS). A total of 329,294 MRS spectra from LAMOST DR8 are utilized in the analysis. The $\mathrm{H}α$ activity index ($I_{\mathrm{H}α}$) and the $\mathrm{H}α$ $R$-index ($R_{\mathrm{H}α}$) are evaluated for the MRS spectra. The $\mathrm{H}α$ chromospheric activity distributions with individual stellar parameters as well as in the $T_\mathrm{eff}$ -- $\log\,g$ and $T_\mathrm{eff}$ -- $\mathrm{[Fe/H]}$ parameter spaces are analyzed based on the $R_{\mathrm{H}α}$ index data. It is found that: (1) for the main-sequence sample, the $R_{\mathrm{H}α}$ distribution with $T_\mathrm{eff}$ has a bowl-shaped lower envelope with a minimum at about 6200 K, a hill-shaped middle envelope with a maximum at about 5600 K, and an upper envelope continuing to increase from hotter to cooler stars; (2) for the giant sample, the middle and upper envelopes of the $R_{\mathrm{H}α}$ distribution first increase with a decrease of $T_\mathrm{eff}$ and then drop to a lower activity level at about 4300 K, revealing different activity characteristics at different stages of stellar evolution; (3) for both the main-sequence and giant samples, the upper envelope of the $R_{\mathrm{H}α}$ distribution with metallicity is higher for stars with $\mathrm{[Fe/H]}$ greater than about $-1.0$, and the lowest-metallicity stars hardly exhibit high $\mathrm{H}α$ indices. A dataset of $\mathrm{H}α$ activity indices for the LAMOST MRS spectra analyzed is provided with this paper.

astro-ph.SR↗
arXiv2023

Towards Optimal Reconstruction of Shear Field with PDF-Folding

Weak lensing provides a direct way of mapping the density distribution in the universe. To reconstruct the density field from the shear catalog, an important step is to build the shear field from the shear catalog, which can be quite nontrivial due to the inhomogeneity of the background galaxy distribution and the shape noise. We propose the PDF-Folding method as a statistically optimal way of reconstructing the shear field. It is an extention of the PDF-SYM method, which is previously designed for optimizing the stacked shear signal as well as the shear-shear correlation for the Fourier\_Quad shear estimators. PDF-Folding does not require smoothing kernels as in traditional methods, therefore it suffers less information loss on small scales, and avoids possible biases due to the spatial variation of shear on the scale of the kernel. We show with analytic reasoning as well as numerical examples that the new method can reach the optimal signal-to-noise ratio on the reconstructed shear map under general observing conditions, i.e., with inhomogeneous background densities or masks. We also show the performance of the new method on real data around foreground galaxy clusters.

astro-ph.CO↗
arXiv2023

Remark on non-contractible closed geodesics and homotopy groups

We prove that if the $m$-th homotopy group for $m \geq 2$ of a closed manifold has non-trivial invariants or coinvariants under the action of the fundamental group, then there exist infinitely many geometrically distinct closed geodesics for a $C^4$-generic Riemannian metric. If moreover there are infinitely many conjugacy classes in the fundamental group, then the same holds for every Riemannian metric.

math.DG↗
arXiv2023

iEnhancer-ELM: improve enhancer identification by extracting position-related multiscale contextual information based on enhancer language models

Motivation: Enhancers are important cis-regulatory elements that regulate a wide range of biological functions and enhance the transcription of target genes. Although many feature extraction methods have been proposed to improve the performance of enhancer identification, they cannot learn position-related multiscale contextual information from raw DNA sequences. Results: In this article, we propose a novel enhancer identification method (iEnhancer-ELM) based on BERT-like enhancer language models. iEnhancer-ELM tokenizes DNA sequences with multi-scale k-mers and extracts contextual information of different scale k-mers related with their positions via an multi-head attention mechanism. We first evaluate the performance of different scale k-mers, then ensemble them to improve the performance of enhancer identification. The experimental results on two popular benchmark datasets show that our model outperforms stateof-the-art methods. We further illustrate the interpretability of iEnhancer-ELM. For a case study, we discover 30 enhancer motifs via a 3-mer-based model, where 12 of motifs are verified by STREME and JASPAR, demonstrating our model has a potential ability to unveil the biological mechanism of enhancer. Availability and implementation: The models and associated code are available at https://github.com/chen-bioinfo/iEnhancer-ELM Contact: junjiechen@hit.edu.cn Supplementary information: Supplementary data are available at Bioinformatics Advances online.

q-bio.GN↗
arXiv2023

Supermassive primordial black holes in multiverse: for nano-Hertz gravitational wave and high-redshift JWST galaxies

Recently, observational hints for supermassive black holes have been accumulating, which has inspired ones to wonder: Can primordial black holes (PBHs) be supermassive, in particular with the mass $M\gtrsim 10^{9}M_\odot$? A supercritical bubble (with an inflating baby universe inside it) that nucleated during inflation can develop into a PBH in our observable Universe. Here, we find that when the inflaton slowly passes by a neighboring vacuum, the nucleating rate of supercritical bubbles would inevitably attain a peak, so the mass distribution of multiverse PBHs, and the mass of peak can be up to $M\gtrsim 10^{11}M_\odot$. Thus our mechanism naturally provides a primordial origin of supermassive BHs.

gr-qc↗
arXiv2023

Effect of thermal fluctuations on spectra and predictability in compressible decaying isotropic turbulence

This study investigates the impact of molecular thermal fluctuations on compressible decaying isotropic turbulence using the unified stochastic particle (USP) method, encompassing both two-dimensional (2D) and three-dimensional (3D) scenarios. The findings reveal that the turbulent spectra of velocity and thermodynamic variables follow the wavenumber scaling law of ${k}^{(d-1)}$ for different spatial dimensions $d$ within the high wavenumber range, indicating the impact of thermal fluctuations on small-scale turbulent statistics. With the application of Helmholtz decomposition, it is found that the thermal fluctuation spectra of solenoidal and compressible velocity components (${\vec{u}}_{s}$ and ${\vec{u}}_{c}$) follow an energy ratio of 1:1 for 2D cases, while the ratio changes to 2:1 for 3D cases. Comparisons between 3D turbulent spectra obtained through USP simulations and direct numerical simulations of the Navier-Stokes equations demonstrate that thermal fluctuations dominate the spectra at length scales comparable to the Kolmogorov length scale. Additionally, the effect of thermal fluctuations on the spectrum of ${\vec{u}}_{c}$ is significantly influenced by variations in the turbulent Mach number. We further study the impact of thermal fluctuations on the predictability of turbulence. With initial differences caused by thermal fluctuations, different flow realizations display significant disparities in velocity and thermodynamic fields at larger scales after a certain period of time, which can be characterized by "inverse error cascades". Moreover, the results suggest a strong correlation between the predictabilities of thermodynamic fields and the predictability of ${\vec{u}}_{c}$.

physics.flu-dyn↗
arXiv2023

Ultraviolet photon-counting single-pixel imaging

We demonstrate photon-counting single-pixel imaging in the ultraviolet region. Toward this target, we develop a high-performance compact single-photon detector based on a 4H-SiC single-photon avalanche diode (SPAD), where a tailored readout circuit with active hold-off time is designed to restrain detector noise and operate the SPAD in free-running mode. We use structured illumination to reconstruct 192$\times$192 compressed images at a 4 fps frame rate. To show the superior capability of ultraviolet characteristics, we use our single-pixel imaging system to identify and distinguish different transparent objects under low-intensity irradiation, and image ultraviolet light sources. The results provide a practical solution for general ultraviolet imaging applications.

physics.optics↗
arXiv2023

Towards Large-scale Single-shot Millimeter-wave Imaging for Low-cost Security Inspection

Millimeter-wave (MMW) imaging is emerging as a promising technique for safe security inspection. It achieves a delicate balance between imaging resolution, penetrability and human safety, resulting in higher resolution compared to low-frequency microwave, stronger penetrability compared to visible light, and stronger safety compared to X ray. Despite of recent advance in the last decades, the high cost of requisite large-scale antenna array hinders widespread adoption of MMW imaging in practice. To tackle this challenge, we report a large-scale single-shot MMW imaging framework using sparse antenna array, achieving low-cost but high-fidelity security inspection under an interpretable learning scheme. We first collected extensive full-sampled MMW echoes to study the statistical ranking of each element in the large-scale array. These elements are then sampled based on the ranking, building the experimentally optimal sparse sampling strategy that reduces the cost of antenna array by up to one order of magnitude. Additionally, we derived an untrained interpretable learning scheme, which realizes robust and accurate image reconstruction from sparsely sampled echoes. Last, we developed a neural network for automatic object detection, and experimentally demonstrated successful detection of concealed centimeter-sized targets using 10% sparse array, whereas all the other contemporary approaches failed at the same sample sampling ratio. The performance of the reported technique presents higher than 50% superiority over the existing MMW imaging schemes on various metrics including precision, recall, and mAP50. With such strong detection ability and order-of-magnitude cost reduction, we anticipate that this technique provides a practical way for large-scale single-shot MMW imaging, and could advocate its further practical applications.

eess.IV↗
arXiv2023

Novel deep learning methods for 3D flow field segmentation and classification

Flow field segmentation and classification help researchers to understand vortex structure and thus turbulent flow. Existing deep learning methods mainly based on global information and focused on 2D circumstance. Based on flow field theory, we propose novel flow field segmentation and classification deep learning methods in three-dimensional space. We construct segmentation criterion based on local velocity information and classification criterion based on the relationship between local vorticity and vortex wake, to identify vortex structure in 3D flow field, and further classify the type of vortex wakes accurately and rapidly. Simulation experiment results showed that, compared with existing methods, our segmentation method can identify the vortex area more accurately, while the time consumption is reduced more than 50%; our classification method can reduce the time consumption by more than 90% while maintaining the same classification accuracy level.

cs.CV↗
arXiv2023

Local laser heating effects in diamond probed by photoluminescence of SiV centers at low temperature

Diamond is generally considered to have high thermal conductivity, so little attention has been paid to the laser heating effects at low excitation power. However, defects during the growth process can result in a great degradation of thermal conductivity, especially at low temperatures. Here, we observed the dynamic redshift and broadening of zero phonon line (ZPL) of silicon-vacancy (SiV) centers in diamondin the experiment. Utilizing the intrinsic temperature response of the fine structure spectra of SiV as a probe, we confirmed that the laser heating effect appears and the temperature rising results from high defect concentration. By simulating the thermal diffusion process, we have estimated the thermal conductivity of around 1 W/(mK) at the local site, which is a two order magnitude lower than that of single-crystal diamond. Our results provide a feasible scheme for characterizing the laser heating effect of diamond at low temperatures.

physics.optics↗
arXiv2023

Language-specific Acoustic Boundary Learning for Mandarin-English Code-switching Speech Recognition

Code-switching speech recognition (CSSR) transcribes speech that switches between multiple languages or dialects within a single sentence. The main challenge in this task is that different languages often have similar pronunciations, making it difficult for models to distinguish between them. In this paper, we propose a method for solving the CSSR task from the perspective of language-specific acoustic boundary learning. We introduce language-specific weight estimators (LSWE) to model acoustic boundary learning in different languages separately. Additionally, a non-autoregressive (NAR) decoder and a language change detection (LCD) module are employed to assist in training. Evaluated on the SEAME corpus, our method achieves a state-of-the-art mixed error rate (MER) of 16.29% and 22.81% on the test_man and test_sge sets. We also demonstrate the effectiveness of our method on a 9000-hour in-house meeting code-switching dataset, where our method achieves a relatively 7.9% MER reduction.

cs.SD↗