arXiv subjects

Chun Li

Publications and source records attributed to Chun Li.

At least 37 records · Page 2Linked to original sources

arXiv2025

A DeepONet joint Neural Tangent Kernel Hybrid Framework for Physics-Informed Inverse Source Problems and Robust Image Reconstruction

This work presents a novel hybrid approach that integrates Deep Operator Networks (DeepONet) with the Neural Tangent Kernel (NTK) to solve complex inverse problem. The method effectively addresses tasks such as source localization governed by the Navier-Stokes equations and image reconstruction, overcoming challenges related to nonlinearity, sparsity, and noisy data. By incorporating physics-informed constraints and task-specific regularization into the loss function, the framework ensures solutions that are both physically consistent and accurate. Validation on diverse synthetic and real datasets demonstrates its robustness, scalability, and precision, showcasing its broad potential applications in computational physics and imaging sciences.

cs.CV↗

arXiv2025

Federated Dialogue-Semantic Diffusion for Emotion Recognition under Incomplete Modalities

Multimodal Emotion Recognition in Conversations (MERC) enhances emotional understanding through the fusion of multimodal signals. However, unpredictable modality absence in real-world scenarios significantly degrades the performance of existing methods. Conventional missing-modality recovery approaches, which depend on training with complete multimodal data, often suffer from semantic distortion under extreme data distributions, such as fixed-modality absence. To address this, we propose the Federated Dialogue-guided and Semantic-Consistent Diffusion (FedDISC) framework, pioneering the integration of federated learning into missing-modality recovery. By federated aggregation of modality-specific diffusion models trained on clients and broadcasting them to clients missing corresponding modalities, FedDISC overcomes single-client reliance on modality completeness. Additionally, the DISC-Diffusion module ensures consistency in context, speaker identity, and semantics between recovered and available modalities, using a Dialogue Graph Network to capture conversational dependencies and a Semantic Conditioning Network to enforce semantic alignment. We further introduce a novel Alternating Frozen Aggregation strategy, which cyclically freezes recovery and classifier modules to facilitate collaborative optimization. Extensive experiments on the IEMOCAP, CMUMOSI, and CMUMOSEI datasets demonstrate that FedDISC achieves superior emotion classification performance across diverse missing modality patterns, outperforming existing approaches.

cs.CV↗

arXiv2025

Diffusion Low Rank Hybrid Reconstruction for Sparse View Medical Imaging

This work presents TV-LoRA, a novel method for low-dose sparse-view CT reconstruction that combines a diffusion generative prior (NCSN++ with SDE modeling) and multi-regularization constraints, including anisotropic TV and nuclear norm (LoRA), within an ADMM framework. To address ill-posedness and texture loss under extremely sparse views, TV-LoRA integrates generative and physical constraints, and utilizes a 2D slice-based strategy with FFT acceleration and tensor-parallel optimization for efficient inference. Experiments on AAPM-2016, CTHD, and LIDC datasets with $N_{\mathrm{view}}=8,4,2$ show that TV-LoRA consistently surpasses benchmarks in SSIM, texture recovery, edge clarity, and artifact suppression, demonstrating strong robustness and generalizability. Ablation studies confirm the complementary effects of LoRA regularization and diffusion priors, while the FFT-PCG module provides a speedup. Overall, Diffusion + TV-LoRA achieves high-fidelity, efficient 3D CT reconstruction and broad clinical applicability in low-dose, sparse-sampling scenarios.

cs.CV↗

arXiv2025

The Stellar Abundances and Galactic Evolution Survey (SAGES). IV. Surface Gravity Estimation and Giant-Dwarf Separation with the DDO51 Filter

Reliable estimation of stellar surface gravity (log $g$) for a large sample is crucial for evaluating stellar evolution models and understanding galactic structure; However, it is not easy to accomplish due to the difficulty in gathering a large spectroscopic data set. Photometric sky survey using a specific filter, on the other hand, can play a substantial role in the assessment of log $g$. The Stellar Abundances and Galactic Evolution Survey (SAGES) utilizes eight filters to provide accurate stellar parameters for $\sim10^{7}$ stars, with its DDO51 intermediate-band filter specifically designed for robust log $g$ determination. In this work, the observed SAGES $u_{\rm SC}$ and $v_{\rm SAGES}$ photometry, the synthetic photometry in $g$, $r$, $i$, and DDO51 bands derived from \textit{Gaia} XP spectra are employed to investigate the importance of the DDO51 filter in the determination of log $g$. We applied machine-learning-based extinction correction and employed XGBoost models, trained on stellar parameters from LAMOST, to predict log $g$ using photometric data. By comparing model predicted log $g$ with LAMOST values, we find that including DDO51 filter improve the accuracies of log $g$ estimates by 21.0\% (from 0.224\,dex to 0.177\,dex) overall, and by 26.5\% (from 0.302\,dex to 0.222\,dex ) for GK-type stars, as compared to those obtained without DDO51. The DDO51 filter is also validated to be particularly effective for metal-poor stars ([Fe/H]$<$-1.0), where it significantly mitigates systematic biases. Our findings highlight the diagnostic power of the SAGES DDO51 filter, providing enhanced stellar characterization vital for future in-depth studies of the Milky Way.

astro-ph.SR↗

arXiv2025

EFTViT: Efficient Federated Training of Vision Transformers with Masked Images on Resource-Constrained Clients

Federated learning research has recently shifted from Convolutional Neural Networks (CNNs) to Vision Transformers (ViTs) due to their superior capacity. ViTs training demands higher computational resources due to the lack of 2D inductive biases inherent in CNNs. However, efficient federated training of ViTs on resource-constrained edge devices remains unexplored in the community. In this paper, we propose EFTViT, a hierarchical federated framework that leverages masked images to enable efficient, full-parameter training on resource-constrained edge devices, offering substantial benefits for learning on heterogeneous data. In general, we patchify images and randomly mask a portion of the patches, observing that excluding them from training has minimal impact on performance while substantially reducing computation costs and enhancing data content privacy protection. Specifically, EFTViT comprises a series of lightweight local modules and a larger global module, updated independently on clients and the central server, respectively. The local modules are trained on masked image patches, while the global module is trained on intermediate patch features uploaded from the local client, balanced through a proposed median sampling strategy to erase client data distribution privacy. We analyze the computational complexity and privacy protection of EFTViT. Extensive experiments on popular benchmarks show that EFTViT achieves up to 28.17% accuracy improvement, reduces local training computational cost by up to 2.8$\times$, and cuts local training time by up to 4.4$\times$ compared to existing methods.

cs.CV↗

arXiv2025

FedVLA: Federated Vision-Language-Action Learning with Dual Gating Mixture-of-Experts for Robotic Manipulation

Vision-language-action (VLA) models have significantly advanced robotic manipulation by enabling robots to interpret language instructions for task execution. However, training these models often relies on large-scale user-specific data, raising concerns about privacy and security, which in turn limits their broader adoption. To address this, we propose FedVLA, the first federated VLA learning framework, enabling distributed model training that preserves data privacy without compromising performance. Our framework integrates task-aware representation learning, adaptive expert selection, and expert-driven federated aggregation, enabling efficient and privacy-preserving training of VLA models. Specifically, we introduce an Instruction Oriented Scene-Parsing mechanism, which decomposes and enhances object-level features based on task instructions, improving contextual understanding. To effectively learn diverse task patterns, we design a Dual Gating Mixture-of-Experts (DGMoE) mechanism, where not only input tokens but also self-aware experts adaptively decide their activation. Finally, we propose an Expert-Driven Aggregation strategy at the federated server, where model aggregation is guided by activated experts, ensuring effective cross-client knowledge transfer.Extensive simulations and real-world robotic experiments demonstrate the effectiveness of our proposals. Notably, DGMoE significantly improves computational efficiency compared to its vanilla counterpart, while FedVLA achieves task success rates comparable to centralized training, effectively preserving data privacy.

cs.RO↗

arXiv2025

Uncertainty Quantification for Incomplete Multi-View Data Using Divergence Measures

Existing multi-view classification and clustering methods typically improve task accuracy by leveraging and fusing information from different views. However, ensuring the reliability of multi-view integration and final decisions is crucial, particularly when dealing with noisy or corrupted data. Current methods often rely on Kullback-Leibler (KL) divergence to estimate uncertainty of network predictions, ignoring domain gaps between different modalities. To address this issue, KPHD-Net, based on Hölder divergence, is proposed for multi-view classification and clustering tasks. Generally, our KPHD-Net employs a variational Dirichlet distribution to represent class probability distributions, models evidences from different views, and then integrates it with Dempster-Shafer evidence theory (DST) to improve uncertainty estimation effects. Our theoretical analysis demonstrates that Proper Hölder divergence offers a more effective measure of distribution discrepancies, ensuring enhanced performance in multi-view learning. Moreover, Dempster-Shafer evidence theory, recognized for its superior performance in multi-view fusion tasks, is introduced and combined with the Kalman filter to provide future state estimations. This integration further enhances the reliability of the final fusion results. Extensive experiments show that the proposed KPHD-Net outperforms the current state-of-the-art methods in both classification and clustering tasks regarding accuracy, robustness, and reliability, with theoretical guarantees.

cs.CV↗

arXiv2025

Strongly Magnetically Tuned Coupling Strength and Nonlinearity in CrSBr Exciton-Polaritons

Two-dimensional van der Waals (vdW) magnetic semiconductors CrSBr offer an ideal platform to achieve exciton-polaritons correlated with magnetic orders for developing solid-state quantum, spintronic, and photonic devices. However, for the exciton-polaritons formed by lower-energy excitons (XL = 1.37 eV), the coupling strength and nonlinear optical response are almost inert to the external magnetic field. Here, we demonstrate robust strong coupling between higher-energy excitons (XH = 1.8 eV) and photons that persists up to room temperature, along with giant magnetic-field tunability. The Rabi splitting energy is tuned up to 100 meV within a moderate 0.45 T in-plane magnetic field due to changes in excitonic states during the spin transitions. Besides, we observe significantly enhanced polariton nonlinearity in the intermediate magnetic phase, which exhibits a distinct mode-number dependence and originates from magnon-assisted long-range attractive interactions and coupling strength reduction. These results advance the development of on-demand polariton platforms for spin-correlated quantum optoelectronics.

cond-mat.mtrl-sci↗

arXiv2025

Robust Brain Tumor Segmentation with Incomplete MRI Modalities Using Hölder Divergence and Mutual Information-Enhanced Knowledge Transfer

Multimodal MRI provides critical complementary information for accurate brain tumor segmentation. However, conventional methods struggle when certain modalities are missing due to issues such as image quality, protocol inconsistencies, patient allergies, or financial constraints. To address this, we propose a robust single-modality parallel processing framework that achieves high segmentation accuracy even with incomplete modalities. Leveraging Holder divergence and mutual information, our model maintains modality-specific features while dynamically adjusting network parameters based on the available inputs. By using these divergence- and information-based loss functions, the framework effectively quantifies discrepancies between predictions and ground-truth labels, resulting in consistently accurate segmentation. Extensive evaluations on the BraTS 2018 and BraTS 2020 datasets demonstrate superior performance over existing methods in handling missing modalities.

cs.CV↗

arXiv2025

Estimating treatment effects with a unified semi-parametric difference-in-differences approach

Difference-in-differences (DID) approaches are widely used for estimating causal effects with observational data before and after an intervention. DID traditionally estimates the average treatment effect among the treated after making a parallel trends assumption on the means of the outcome. With skewed outcomes, a transformation is often needed; however, the transformation may be difficult to choose, results may be sensitive to the choice, and parallel trends assumptions are made on the transformed scale. Recent DID methods estimate alternative treatment effects that may be preferable with skewed outcomes. However, each alternative DID estimator requires a different parallel trends assumption. We introduce a new DID method capable of estimating average, quantile, probability, and novel Mann-Whitney treatment effects among the treated with a single unifying parallel trends assumption. The proposed method uses a semi-parametric cumulative probability model (CPM). The CPM is a linear model for a latent variable on covariates, where the latent variable results from an unspecified transformation of the outcome. Our DID approach makes a universal parallel trends assumption on the expectation of the latent variable conditional on covariates. Hence, our method avoids specifying outcome transformations and does not require separate assumptions for each estimand. We introduce the method; describe identification, estimation, and inference; conduct simulations evaluating its performance; and apply it to assess the impact of Medicaid expansion on CD4 count among people with HIV.

stat.ME↗

arXiv2025

DeepRAG: Integrating Hierarchical Reasoning and Process Supervision for Biomedical Multi-Hop QA

We propose DeepRAG, a novel framework that integrates DeepSeek hierarchical question decomposition capabilities with RAG Gym unified retrieval-augmented generation optimization using process level supervision. Targeting the challenging MedHopQA biomedical question answering task, DeepRAG systematically decomposes complex queries into precise sub-queries and employs concept level reward signals informed by the UMLS ontology to enhance biomedical accuracy. Preliminary evaluations on the MedHopQA dataset indicate that DeepRAG significantly outperforms baseline models, including standalone DeepSeek and RAG Gym, achieving notable improvements in both Exact Match and concept level accuracy.

cs.CL↗

arXiv2025

Uncertainty Quantification via Hölder Divergence for Multi-View Representation Learning

Evidence-based deep learning represents a burgeoning paradigm for uncertainty estimation, offering reliable predictions with negligible extra computational overheads. Existing methods usually adopt Kullback-Leibler divergence to estimate the uncertainty of network predictions, ignoring domain gaps among various modalities. To tackle this issue, this paper introduces a novel algorithm based on Hölder Divergence (HD) to enhance the reliability of multi-view learning by addressing inherent uncertainty challenges from incomplete or noisy data. Generally, our method extracts the representations of multiple modalities through parallel network branches, and then employs HD to estimate the prediction uncertainties. Through the Dempster-Shafer theory, integration of uncertainty from different modalities, thereby generating a comprehensive result that considers all available representations. Mathematically, HD proves to better measure the ``distance'' between real data distribution and predictive distribution of the model and improve the performances of multi-class recognition tasks. Specifically, our method surpass the existing state-of-the-art counterparts on all evaluating benchmarks. We further conduct extensive experiments on different backbones to verify our superior robustness. It is demonstrated that our method successfully pushes the corresponding performance boundaries. Finally, we perform experiments on more challenging scenarios, \textit{i.e.}, learning with incomplete or noisy data, revealing that our method exhibits a high tolerance to such corrupted data.

cs.CV↗

arXiv2025

The Stellar Abundances and Galactic Evolution Survey (SAGES). II. Machine Learning-Based Stellar parameters for 21 million stars from the First Data Release

Stellar parameters for large samples of stars play a crucial role in constraining the nature of stars and stellar populations in the Galaxy. An increasing number of medium-band photometric surveys are presently used in estimating stellar parameters. In this study, we present a machine-learning approach to derive estimates of stellar parameters, including [Fe/H], logg, and Teff, based on a combination of medium-band and broad-band photometric observations. Our analysis employs data primarily sourced from the SAGE Survey , which aims to observe much of the Northern Hemisphere. We combine the $uv$-band data from SAGES DR1 with photometric and astrometric data from Gaia EDR3, and apply the random forest method to estimate stellar parameters for approximately 21 million stars. We are able to obtain precisions of 0.09 dex for [Fe/H], 0.12 dex for logg, and 70 K for Teff. Furthermore, by incorporating 2MASS and WISE infrared photometric and GALEX ultraviolet data, we are able to achieve even higher precision estimates for over 2.2 million stars. These results are applicable to both giant and dwarf stars. Building upon this mapping, we construct a foundational dataset for research on metal-poor stars, the structure of the Milky Way, and beyond. With the forthcoming release of additional bands from SAGE Survey such DDO51 and H-alpha, this versatile machine learning approach is poised to play an important role in upcoming surveys featuring expanded filter sets

astro-ph.SR↗

arXiv2025

Between- and Within-Cluster Spearman Rank Correlations

Clustered data are common in practice. Clustering arises when subjects are measured repeatedly, or subjects are nested in groups (e.g., households, schools). It is often of interest to evaluate the correlation between two variables with clustered data. There are three commonly used Pearson correlation coefficients (total, between-, and within-cluster), which together provide an enriched perspective of the correlation. However, these Pearson correlation coefficients are sensitive to extreme values and skewed distributions. They also vary with data transformation, which is arbitrary and often difficult to choose, and they are not applicable to ordered categorical data. Current nonparametric correlation measures for clustered data are only for the total correlation. Here we define population parameters for the between- and within-cluster Spearman rank correlations. The definitions are natural extensions of the Pearson between- and within-cluster correlations to the rank scale. We show that the total Spearman rank correlation approximates a linear combination of the between- and within-cluster Spearman rank correlations, where the weights are functions of rank intraclass correlations of the two random variables. We also discuss the equivalence between the within-cluster Spearman rank correlation and the covariate-adjusted partial Spearman rank correlation. Furthermore, we describe estimation and inference for the three Spearman rank correlations, conduct simulations to evaluate the performance of our estimators, and illustrate their use with data from a longitudinal biomarker study and a clustered randomized trial.

stat.ME↗

arXiv2025

General Information Metrics for Improving AI Model Training Efficiency

To address the growing size of AI model training data and the lack of a universal data selection methodology-factors that significantly drive up training costs -- this paper presents the General Information Metrics Evaluation (GIME) method. GIME leverages general information metrics from Objective Information Theory (OIT), including volume, delay, scope, granularity, variety, duration, sampling rate, aggregation, coverage, distortion, and mismatch to optimize dataset selection for training purposes. Comprehensive experiments conducted across diverse domains, such as CTR Prediction, Civil Case Prediction, and Weather Forecasting, demonstrate that GIME effectively preserves model performance while substantially reducing both training time and costs. Additionally, applying GIME within the Judicial AI Program led to a remarkable 39.56% reduction in total model training expenses, underscoring its potential to support efficient and sustainable AI development.

cs.LG↗

arXiv2024

Interharmonic Power: A New Concept for Power System Oscillation Source Location

Power system oscillations are a significant concern for system operators, a problem that has grown due to the interconnection of inverter-based resources. To address this issue, various methods have been proposed to locate the sources of oscillations, which is essential for effective mitigation actions. A common characteristic of these methods is that they rely on phasor representation of oscillation phenomena. This paper takes a different approach by examining the actual voltage and current waveforms underlying the phasors. It is found that the presence of interharmonic components is both the necessary and sufficient condition for phasor oscillations. Moreover, the generation and propagation of interharmonic powers are identified as the true culprits behind power system oscillations and oscillatory instability. Based on these insights, two new methods are developed for locating oscillation sources: one for measurement-based monitoring applications and another for model-based system studies. These findings are validated through four field data-based and one simulation-based case studies.

eess.SY↗

arXiv2024

Robust Divergence Learning for Missing-Modality Segmentation

Multimodal Magnetic Resonance Imaging (MRI) provides essential complementary information for analyzing brain tumor subregions. While methods using four common MRI modalities for automatic segmentation have shown success, they often face challenges with missing modalities due to image quality issues, inconsistent protocols, allergic reactions, or cost factors. Thus, developing a segmentation paradigm that handles missing modalities is clinically valuable. A novel single-modality parallel processing network framework based on Hölder divergence and mutual information is introduced. Each modality is independently input into a shared network backbone for parallel processing, preserving unique information. Additionally, a dynamic sharing framework is introduced that adjusts network parameters based on modality availability. A Hölder divergence and mutual information-based loss functions are used for evaluating discrepancies between predictions and labels. Extensive testing on the BraTS 2018 and BraTS 2020 datasets demonstrates that our method outperforms existing techniques in handling missing modalities and validates each component's effectiveness.

eess.IV↗

arXiv2024

Detector integration at HEPS: a systematic, efficient and high-performance approach

At least 25 kinds of detector-like devices need to be integrated in Phase I of the High Energy Photon Source (HEPS), and the work needs to be carefully planned to maximise productivity with highly limited human resources. After a systematic analysis on the actual work involved in detector integration, a separation of concerns between collaborating groups of personnel is established to minimise the duplication of efforts. To facilitate software development for detector integration, the ADGenICam library, which abstracts repeated code in EPICS modules for cameras, is extended to support a much wider range of detectors. An increasingly considerable fraction of detectors, both inside and outside HEPS, offer performance that exceed capabilities of the areaDetector framework in EPICS. Given this background, areaDetector's limitations in performance and architecture are analysed, and a QueueIOC -based framework that overcomes these limitations is introduced. A simple, flexible ZeroMQ-based protocol is used for data transport in this framework, while RDMA transport and multi-node readout will be explored for higher data throughputs. By calling C/C++ libraries from within Python, the performance of the former and the expressiveness of the latter can coexist nicely; the expressiveness allows for much higher efficiency in the implementation and use of integration modules functionally comparable to their EPICS counterparts.

physics.ins-det↗