arXiv subjects

Jia Zhang

Publications and source records attributed to Jia Zhang.

At least 55 records · Page 3Linked to original sources

arXiv2025

The bimodal distribution of donor stars in X-ray binaries

The classification of X-ray binaries into high- and low-mass types has historically lacked a unified, data-driven quantitative criterion, and large-scale statistical studies of the donor star population have been limited. In this work, we address this gap by compiling data for 3,964 XRBs and deriving a plentiful set of physical parameters (mass, radius, age, and evolutionary stage) for a sub-sample of 288 donor stars using Gaia DR3 spectral data and stellar evolution models. We find a statistically bimodal distribution in the donor star parameters, which is characterized by a valley at approximately 3 $M_{\odot}$ or 11,000 K. We uncover the physical mechanism behind this bimodality: a previously unreported ``parallel tracks'' phenomenon observed in the relationship between the donor's evolutionary stage and its fundamental parameters, such as luminosity and radius. These two tracks represent distinct main-sequence populations, and the valley between them corresponds to the sparsely populated pre- and post-main-sequence evolutionary phases.

astro-ph.SR↗

arXiv2025

OralGPT: A Two-Stage Vision-Language Model for Oral Mucosal Disease Diagnosis and Description

Oral mucosal diseases such as leukoplakia, oral lichen planus, and recurrent aphthous ulcers exhibit diverse and overlapping visual features, making diagnosis challenging for non-specialists. While vision-language models (VLMs) have shown promise in medical image interpretation, their application in oral healthcare remains underexplored due to the lack of large-scale, well-annotated datasets. In this work, we present \textbf{OralGPT}, the first domain-specific two-stage vision-language framework designed for oral mucosal disease diagnosis and captioning. In Stage 1, OralGPT learns visual representations and disease-related concepts from classification labels. In Stage 2, it enhances its language generation ability using long-form expert-authored captions. To overcome the annotation bottleneck, we propose a novel similarity-guided data augmentation strategy that propagates descriptive knowledge from expert-labeled images to weakly labeled ones. We also construct the first benchmark dataset for oral mucosal diseases, integrating multi-source image data with both structured and unstructured textual annotations. Experimental results on four common oral conditions demonstrate that OralGPT achieves competitive diagnostic performance while generating fluent, clinically meaningful image descriptions. This study provides a foundation for language-assisted diagnostic tools in oral healthcare.

q-bio.QM↗

arXiv2025

Automated Glaucoma Report Generation via Dual-Attention Semantic Parallel-LSTM and Multimodal Clinical Data Integration

Generative AI for automated glaucoma diagnostic report generation faces two predominant challenges: content redundancy in narrative outputs and inadequate highlighting of pathologically significant features including optic disc cupping, retinal nerve fiber layer defects, and visual field abnormalities. These limitations primarily stem from current multimodal architectures' insufficient capacity to extract discriminative structural-textural patterns from fundus imaging data while maintaining precise semantic alignment with domain-specific terminology in comprehensive clinical reports. To overcome these constraints, we present the Dual-Attention Semantic Parallel-LSTM Network (DA-SPL), an advanced multimodal generation framework that synergistically processes both fundus imaging and supplementary visual inputs. DA-SPL employs an Encoder-Decoder structure augmented with the novel joint dual-attention mechanism in the encoder for cross-modal feature refinement, the parallelized LSTM decoder architecture for enhanced temporal-semantic consistency, and the specialized label enhancement module for accurate disease-relevant term generation. Rigorous evaluation on standard glaucoma datasets demonstrates DA-SPL's consistent superiority over state-of-the-art models across quantitative metrics. DA-SPL exhibits exceptional capability in extracting subtle pathological indicators from multimodal inputs while generating diagnostically precise reports that exhibit strong concordance with clinical expert annotations.

cs.CE↗

arXiv2025

Longitudinal transport spin polarization of spin degenerate antiferromagnets

A vital goal in spintronics is the efficient electrical generation of spin currents, a pursuit that has recently focused on using antiferromagnets (AFMs) as spin current sources. It has been demonstrated that antiferromagnets with broken PT symmetry (parity + time reversal) can efficiently generate longitudinal and transverse spin currents. At the same time, it has been generally thought that antiferromagnets with PT symmetry (PT-AFMs) forbid the longitudinal spin polarization due to their spin-degenerate band structure. Here, in contrast to this common expectation, we show, using theoretical analysis based on magnetic point group symmetry, that most PT-AFMs can generate longitudinal spin currents due to spin-orbit coupling. Using density-functional theory, we calculate the longitudinal spin conductivity of representative PT-AFMs, L10-MnPt and Mn2Au, and show that its magnitude is comparable to that of their PT-broken counterparts. Our symmetry-enabled classification of antiferromagnets and theoretical results for the longitudinal spin conductivity in representative PT-AFMs expands our understanding of spin transport and shows the possibility of robust spin-current generation in a broad range of spin-degenerate antiferromagnets.

cond-mat.mtrl-sci↗

arXiv2025

Scalable Audio-Visual Masked Autoencoders for Efficient Affective Video Facial Analysis

Affective video facial analysis (AVFA) has emerged as a key research field for building emotion-aware intelligent systems, yet this field continues to suffer from limited data availability. In recent years, the self-supervised learning (SSL) technique of Masked Autoencoders (MAE) has gained momentum, with growing adaptations in its audio-visual contexts. While scaling has proven essential for breakthroughs in general multi-modal learning domains, its specific impact on AVFA remains largely unexplored. Another core challenge in this field is capturing both intra- and inter-modal correlations through scalable audio-visual representations. To tackle these issues, we propose AVF-MAE++, a family of audio-visual MAE models designed to efficiently investigate the scaling properties in AVFA while enhancing cross-modal correlation modeling. Our framework introduces a novel dual masking strategy across audio and visual modalities and strengthens modality encoders with a more holistic design to better support scalable pre-training. Additionally, we present the Iterative Audio-Visual Correlation Learning Module, which improves correlation learning within the SSL paradigm, bridging the limitations of previous methods. To support smooth adaptation and reduce overfitting risks, we further introduce a progressive semantic injection strategy, organizing the model training into three structured stages. Extensive experiments conducted on 17 datasets, covering three major AVFA tasks, demonstrate that AVF-MAE++ achieves consistent state-of-the-art performance across multiple benchmarks. Comprehensive ablation studies further highlight the importance of each proposed component and provide deeper insights into the design choices driving these improvements. Our code and models have been publicly released at Github.

cs.CV↗

arXiv2025

GlaLSTM: A Concurrent LSTM Stream Framework for Glaucoma Detection via Biomarker Mining

Glaucoma is a complex group of eye diseases marked by optic nerve damage, commonly linked to elevated intraocular pressure and biomarkers like retinal nerve fiber layer thickness. Understanding how these biomarkers interact is crucial for unraveling glaucoma's underlying mechanisms. In this paper, we propose GlaLSTM, a novel concurrent LSTM stream framework for glaucoma detection, leveraging latent biomarker relationships. Unlike traditional CNN-based models that primarily detect glaucoma from images, GlaLSTM provides deeper interpretability, revealing the key contributing factors and enhancing model transparency. This approach not only improves detection accuracy but also empowers clinicians with actionable insights, facilitating more informed decision-making. Experimental evaluations confirm that GlaLSTM surpasses existing state-of-the-art methods, demonstrating its potential for both advanced biomarker analysis and reliable glaucoma detection.

eess.IV↗

arXiv2025

X-GAN: A Generative AI-Powered Unsupervised Model for Main Vessel Segmentation of Glaucoma Screening

Structural changes in main retinal blood vessels serve as critical biomarkers for the onset and progression of glaucoma. Identifying these vessels is vital for vascular modeling yet highly challenging. This paper proposes X-GAN, a generative AI-powered unsupervised segmentation model designed for extracting main blood vessels from Optical Coherence Tomography Angiography (OCTA) images. The process begins with the Space Colonization Algorithm (SCA) to rapidly generate a skeleton of vessels, featuring their radii. By synergistically integrating the generative adversarial network (GAN) with biostatistical modeling of vessel radii, X-GAN enables a fast reconstruction of both 2D and 3D representations of the vessels. Based on this reconstruction, X-GAN achieves nearly 100\% segmentation accuracy without relying on labeled data or high-performance computing resources. Experimental results confirm X-GAN's superiority in evaluating main vessel segmentation compared to existing deep learning models.

eess.IV↗

arXiv2025

UniGenX: a unified generative foundation model that couples sequence, structure and function to accelerate scientific design across proteins, molecules and materials

Function in natural systems arises from one-dimensional sequences forming three-dimensional structures with specific properties. However, current generative models suffer from critical limitations: training objectives seldom target function directly, discrete sequences and continuous coordinates are optimized in isolation, and conformational ensembles are under-modeled. We present UniGenX, a unified generative foundation model that addresses these gaps by co-generating sequences and coordinates under direct functional and property objectives across proteins, molecules, and materials. UniGenX represents heterogeneous inputs as a mixed stream of symbolic and numeric tokens, where a decoder-only autoregressive transformer provides global context and a conditional diffusion head generates numeric fields steered by task-specific tokens. Besides the new high SOTAs on structure prediction tasks, the model demonstrates state-of-the-art or competitive performance for the function-aware generation across domains: in materials, it achieves "conflicted" multi-property conditional generation, yielding 436 crystal candidates meeting triple constraints, including 11 with novel compositions; in chemistry, it sets new benchmarks on five property targets and conformer ensemble generation on GEOM; and in biology, it improves success in modeling protein induced fit (RMSD < 2 Å) by over 23-fold and enhances EC-conditioned enzyme design. Ablation studies and cross-domain transfer substantiate the benefits of joint discrete-continuous training, establishing UniGenX as a significant advance from prediction to controllable, function-aware generation.

cs.LG↗

arXiv2025

eMotions: A Large-Scale Dataset and Audio-Visual Fusion Network for Emotion Analysis in Short-form Videos

Short-form videos (SVs) have become a vital part of our online routine for acquiring and sharing information. Their multimodal complexity poses new challenges for video analysis, highlighting the need for video emotion analysis (VEA) within the community. Given the limited availability of SVs emotion data, we introduce eMotions, a large-scale dataset consisting of 27,996 videos with full-scale annotations. To ensure quality and reduce subjective bias, we emphasize better personnel allocation and propose a multi-stage annotation procedure. Additionally, we provide the category-balanced and test-oriented variants through targeted sampling to meet diverse needs. While there have been significant studies on videos with clear emotional cues (e.g., facial expressions), analyzing emotions in SVs remains a challenging task. The challenge arises from the broader content diversity, which introduces more distinct semantic gaps and complicates the representations learning of emotion-related features. Furthermore, the prevalence of audio-visual co-expressions in SVs leads to the local biases and collective information gaps caused by the inconsistencies in emotional expressions. To tackle this, we propose AV-CANet, an end-to-end audio-visual fusion network that leverages video transformer to capture semantically relevant representations. We further introduce the Local-Global Fusion Module designed to progressively capture the correlations of audio-visual features. Besides, EP-CE Loss is constructed to globally steer optimizations with tripolar penalties. Extensive experiments across three eMotions-related datasets and four public VEA datasets demonstrate the effectiveness of our proposed AV-CANet, while providing broad insights for future research. Moreover, we conduct ablation studies to examine the critical components of our method. Dataset and code will be made available at Github.

cs.CV↗

arXiv2025

HOLA: Enhancing Audio-visual Deepfake Detection via Hierarchical Contextual Aggregations and Efficient Pre-training

Advances in Generative AI have made video-level deepfake detection increasingly challenging, exposing the limitations of current detection techniques. In this paper, we present HOLA, our solution to the Video-Level Deepfake Detection track of 2025 1M-Deepfakes Detection Challenge. Inspired by the success of large-scale pre-training in the general domain, we first scale audio-visual self-supervised pre-training in the multimodal video-level deepfake detection, which leverages our self-built dataset of 1.81M samples, thereby leading to a unified two-stage framework. To be specific, HOLA features an iterative-aware cross-modal learning module for selective audio-visual interactions, hierarchical contextual modeling with gated aggregations under the local-global perspective, and a pyramid-like refiner for scale-aware cross-grained semantic enhancements. Moreover, we propose the pseudo supervised singal injection strategy to further boost model performance. Extensive experiments across expert models and MLLMs impressivly demonstrate the effectiveness of our proposed HOLA. We also conduct a series of ablation studies to explore the crucial design factors of our introduced components. Remarkably, our HOLA ranks 1st, outperforming the second by 0.0476 AUC on the TestA set.

cs.CV↗

arXiv2025

Nature Language Model: Deciphering the Language of Nature for Scientific Discovery

Foundation models have revolutionized natural language processing and artificial intelligence, significantly enhancing how machines comprehend and generate human languages. Inspired by the success of these foundation models, researchers have developed foundation models for individual scientific domains, including small molecules, materials, proteins, DNA, RNA and even cells. However, these models are typically trained in isolation, lacking the ability to integrate across different scientific domains. Recognizing that entities within these domains can all be represented as sequences, which together form the "language of nature", we introduce Nature Language Model (NatureLM), a sequence-based science foundation model designed for scientific discovery. Pre-trained with data from multiple scientific domains, NatureLM offers a unified, versatile model that enables various applications including: (i) generating and optimizing small molecules, proteins, RNA, and materials using text instructions; (ii) cross-domain generation/design, such as protein-to-molecule and protein-to-RNA generation; and (iii) top performance across different domains, matching or surpassing state-of-the-art specialist models. NatureLM offers a promising generalist approach for various scientific tasks, including drug discovery (hit generation/optimization, ADMET optimization, synthesis), novel material design, and the development of therapeutic proteins or nucleotides. We have developed NatureLM models in different sizes (1 billion, 8 billion, and 46.7 billion parameters) and observed a clear improvement in performance as the model size increases.

cs.AI↗

arXiv2025

Photostriction-tunable Polarization and Structural Dynamics in Interlayer Sliding Ferroelectrics

Two-dimensional ferroelectrics with robust polarization offer promising opportunities for non-volatile memory, field-effect transistors, and optoelectronic devices. However, the impact of lattice deformation on polarization and photoinduced structural response remains poorly understood. Here, we employ first-principles calculations to demonstrate photodoping-induced lattice expansion in rhombohedrally stacked bilayer MoS2, revealing a strong coupling between photodoping carrier and lattice structure. We identify a pronounced photostrictive response in sliding ferroelectrics, wherein electron-hole excitation leads to substantial in-plane expansion, increased interlayer spacing, and enhanced ferroelectric polarization. This strain-induced modulation drives significant bandgap renormalization. The photostriction-tunable polarization and structural dynamics arise from the strong electromechanical coupling inherent to the non-centrosymmetric rhombohedral stacking. The findings provide critical insights into the nonthermal lattice expansion governing sliding ferroelectrics at atomic-scale timescales, while simultaneously laying the groundwork for next-generation electronic and memory technologies by leveraging lattice-tunable polarization switching.

cond-mat.mtrl-sci↗

arXiv2025

Efficient and Scalable Density Functional Theory Hamiltonian Prediction through Adaptive Sparsity

Hamiltonian matrix prediction is pivotal in computational chemistry, serving as the foundation for determining a wide range of molecular properties. While SE(3) equivariant graph neural networks have achieved remarkable success in this domain, their substantial computational cost--driven by high-order tensor product (TP) operations--restricts their scalability to large molecular systems with extensive basis sets. To address this challenge, we introduce SPHNet, an efficient and scalable equivariant network, that incorporates adaptive SParsity into Hamiltonian prediction. SPHNet employs two innovative sparse gates to selectively constrain non-critical interaction combinations, significantly reducing tensor product computations while maintaining accuracy. To optimize the sparse representation, we develop a Three-phase Sparsity Scheduler, ensuring stable convergence and achieving high performance at sparsity rates of up to 70%. Extensive evaluations on QH9 and PubchemQH datasets demonstrate that SPHNet achieves state-of-the-art accuracy while providing up to a 7x speedup over existing models. Beyond Hamiltonian prediction, the proposed sparsification techniques also hold significant potential for improving the efficiency and scalability of other SE(3) equivariant networks, further broadening their applicability and impact. Our code can be found at https://github.com/microsoft/SPHNet.

cs.LG↗

arXiv2025

D3: Diversity, Difficulty, and Dependability-Aware Data Selection for Sample-Efficient LLM Instruction Tuning

Recent advancements in instruction tuning for large language models (LLMs) suggest that a small, high-quality dataset can significantly equip LLMs with instruction-following capabilities, outperforming large datasets often burdened by quality and redundancy issues. However, the challenge lies in automatically identifying valuable subsets from large datasets to boost both the effectiveness and efficiency of instruction tuning. In this paper, we first establish data selection criteria based on three distinct aspects of data value: diversity, difficulty, and dependability, and then propose the D3 method comprising two key steps of scoring and selection. Specifically, in the scoring step, we define the diversity function to measure sample distinctiveness and introduce the uncertainty-based prediction difficulty to evaluate sample difficulty by mitigating the interference of context-oriented generation diversity. Additionally, we integrate an external LLM for dependability assessment. In the selection step, we formulate the D3 weighted coreset objective, which jointly optimizes three aspects of data value to solve for the most valuable subset. The two steps of D3 can iterate multiple rounds, incorporating feedback to refine the selection focus adaptively. Experiments on both public datasets and the real-world Taobao Live application demonstrate the effectiveness of D3 in endowing LLMs with competitive or even superior instruction-following capabilities using less than 10\% of the entire dataset.

cs.LG↗

arXiv2025

Investigating the $4D_{3/2}|3,\pm2\rangle$--$4D_{5/2}|3,\pm2\rangle$ transition in Nb$^{4+}$ for a THz atomic clock

In this work, the $4D_{3/2}|3,\pm2\rangle \rightarrow 4D_{5/2}|3,\pm2\rangle$ transition in the Nb$^{4+}$ ion is identified as a promising candidate for a terahertz (THz) atomic clock, with the transition frequency occurring at 56.0224 THz. This transition is primarily driven by the magnetic dipole decay channel, which can easily be accessed by a laser. We focus on the stable $^{93}$Nb isotope, which has 100\% natural abundance and a nuclear spin of $I=9/2$ for experimental advantage. Our data analysis allows us to estimate potential systematic shifts in the proposed clock system, including those due to blackbody radiation, electric quadrupole, second-order Zeeman, and second-order Doppler {shifts}. {The scheme presented in this study can help suppress the AC Stark and electric quadrupole shifts in the clock frequency measurement.} {All these analyses} suggest that the proposed THz atomic clock using Nb$^{4+}$ could be valuable in both quantum thermometry and frequency metrology.

physics.atom-ph↗

arXiv2025

Probability Distribution Alignment and Low-Rank Weight Decomposition for Source-Free Domain Adaptive Brain Decoding

Brain decoding currently faces significant challenges in individual differences, modality alignment, and high-dimensional embeddings. To address individual differences, researchers often use source subject data, which leads to issues such as privacy leakage and heavy data storage burdens. In modality alignment, current works focus on aligning the softmax probability distribution but neglect the alignment of marginal probability distributions, resulting in modality misalignment. Additionally, images and text are aligned separately with fMRI without considering the complex interplay between images and text, leading to poor image reconstruction. Finally, the enormous dimensionality of CLIP embeddings causes significant computational costs. Although the dimensionality of CLIP embeddings can be reduced by ignoring the number of patches obtained from images and the number of tokens acquired from text, this comes at the cost of a significant drop in model performance, creating a dilemma. To overcome these limitations, we propose a source-free domain adaptation-based brain decoding framework.

cs.CV↗

arXiv2025

Leggiero: Analog WiFi Backscatter with Payload Transparency

Backscatter is an enabling technology for battery-free sensing in today's Artificial Intelligence of Things (AIOT). Building a backscatter-based sensing system, however, is a daunting task, due to two obstacles: the unaffordable power consumption of the microprocessor and the coexistence with the ambient carrier's traffic. In order to address the above issues, in this paper, we present Leggiero, the first-of-its-kind analog WiFi backscatter with payload transparency. Leveraging a specially designed circuit with a varactor diode, this design avoids using a microprocessor to interface between the radio and the sensor, and directly converts the analog sensor signal into the phase of RF (radio frequency) signal. By carefully designing the reference circuit on the tag and precisely locating the extra long training field (LTF) section of a WiFi packet, Leggiero embeds the analog phase value into the channel state information (CSI). A commodity WiFi receiver without hardware modification can simultaneously decode the WiFi and the sensor data. We implement Leggiero design and evaluate its performance under varied settings. The results show that the power consumption of the Leggiero tag (excluding the power of the peripheral sensor module) is 30uW at a sampling rate of 400Hz, which is 4.8* and 4* lower than the state-of-the-art WiFi backscatter schemes. The uplink throughput of Leggiero is suficient to support a variety of sensing applications, while keeping the WiFi carrier's throughput performance unaffected.

cs.NI↗

arXiv2025

QuinID: Enabling FDMA-Based Fully Parallel RFID with Frequency-Selective Antenna

Parallelizing passive Radio Frequency Identification (RFID) reading is an arguably crucial, yet unsolved challenge in modern IoT applications. Existing approaches remain limited to time-division operations and fail to read multiple tags simultaneously. In this paper, we introduce QuinID, the first frequency-division multiple access (FDMA) RFID system to achieve fully parallel reading. We innovatively exploit the frequency selectivity of the tag antenna rather than a conventional digital FDMA, bypassing the power and circuitry constraint of RFID tags. Specifically, we delicately design the frequency-selective antenna based on surface acoustic wave (SAW) components to achieve extreme narrow-band response, so that QuinID tags (i.e., QuinTags) operate exclusively within their designated frequency bands. By carefully designing the matching network and canceling various interference, a customized QuinReader communicates simultaneously with multiple QuinTags across distinct bands. QuinID maintains high compatibility with commercial RFID systems and presents a tag cost of less than 10 cents. We implement a 5-band QuinID system and evaluate its performance under various settings. The results demonstrate a fivefold increase in read rate, reaching up to 5000 reads per second.

cs.NI↗