arXiv subjects

Xinyu Zhang

Publications and source records attributed to Xinyu Zhang.

At least 199 records · Page 11Linked to original sources

arXiv2025

Orientation-Aware Sparse Tensor PCA for Efficient Unsupervised Feature Selection

Recently, introducing Tensor Decomposition (TD) techniques into unsupervised feature selection (UFS) has been an emerging research topic. A tensor structure is beneficial for mining the relations between different modes and helps relieve the computation burden. However, while existing methods exploit TD to preserve the data tensor structure, they do not consider the influence of data orientation and thus have difficulty in handling orientation-specific data such as time series. To solve the above problem, we utilize the orientation-dependent tensor-tensor product from Tensor Singular Value Decomposition based on *M-product (T-SVDM) and extend the one-dimensional Sparse Principal Component Analysis (SPCA) to a tensor form. The proposed sparse tensor PCA model can constrain sparsity at the specified mode and yield sparse tensor principal components, enhancing flexibility and accuracy in learning feature relations. To ensure fast convergence and a flexible description of feature correlation, we develop a convex version specially designed for general UFS tasks and propose an efficient slice-by-slice algorithm that performs dual optimization in the transform domain. Experimental results on real-world datasets demonstrate the effectiveness and remarkable computational efficiency of the proposed method for tensor data of diverse structures over the state-of-the-art. When transform axes align with feature distribution patterns, our method is promising for various applications. The codes related to our proposed methods and the experiments are available at https://github.com/zjj20212035/STPCA.git.

cs.LG↗

arXiv2025

HAC-LOCO: Learning Hierarchical Active Compliance Control for Quadruped Locomotion under Continuous External Disturbances

Despite recent remarkable achievements in quadruped control, it remains challenging to ensure robust and compliant locomotion in the presence of unforeseen external disturbances. Existing methods prioritize locomotion robustness over compliance, often leading to stiff, high-frequency motions, and energy inefficiency. This paper, therefore, presents a two-stage hierarchical learning framework that can learn to take active reactions to external force disturbances based on force estimation. In the first stage, a velocity-tracking policy is trained alongside an auto-encoder to distill historical proprioceptive features. A neural network-based estimator is learned through supervised learning, which estimates body velocity and external forces based on proprioceptive measurements. In the second stage, a compliance action module, inspired by impedance control, is learned based on the pre-trained encoder and policy. This module is employed to actively adjust velocity commands in response to external forces based on real-time force estimates. With the compliance action module, a quadruped robot can robustly handle minor disturbances while appropriately yielding to significant forces, thus striking a balance between robustness and compliance. Simulations and real-world experiments have demonstrated that our method has superior performance in terms of robustness, energy efficiency, and safety. Experiment comparison shows that our method outperforms the state-of-the-art RL-based locomotion controllers. Ablation studies are given to show the critical roles of the compliance action module.

cs.RO↗

arXiv2025

Seedream 3.0 Technical Report

We present Seedream 3.0, a high-performance Chinese-English bilingual image generation foundation model. We develop several technical improvements to address existing challenges in Seedream 2.0, including alignment with complicated prompts, fine-grained typography generation, suboptimal visual aesthetics and fidelity, and limited image resolutions. Specifically, the advancements of Seedream 3.0 stem from improvements across the entire pipeline, from data construction to model deployment. At the data stratum, we double the dataset using a defect-aware training paradigm and a dual-axis collaborative data-sampling framework. Furthermore, we adopt several effective techniques such as mixed-resolution training, cross-modality RoPE, representation alignment loss, and resolution-aware timestep sampling in the pre-training phase. During the post-training stage, we utilize diversified aesthetic captions in SFT, and a VLM-based reward model with scaling, thereby achieving outputs that well align with human preferences. Furthermore, Seedream 3.0 pioneers a novel acceleration paradigm. By employing consistent noise expectation and importance-aware timestep sampling, we achieve a 4 to 8 times speedup while maintaining image quality. Seedream 3.0 demonstrates significant improvements over Seedream 2.0: it enhances overall capabilities, in particular for text-rendering in complicated Chinese characters which is important to professional typography generation. In addition, it provides native high-resolution output (up to 2K), allowing it to generate images with high visual quality.

cs.CV↗

arXiv2025

Let Your Video Listen to Your Music!

Aligning the rhythm of visual motion in a video with a given music track is a practical need in multimedia production, yet remains an underexplored task in autonomous video editing. Effective alignment between motion and musical beats enhances viewer engagement and visual appeal, particularly in music videos, promotional content, and cinematic editing. Existing methods typically depend on labor-intensive manual cutting, speed adjustments, or heuristic-based editing techniques to achieve synchronization. While some generative models handle joint video and music generation, they often entangle the two modalities, limiting flexibility in aligning video to music beats while preserving the full visual content. In this paper, we propose a novel and efficient framework, termed MVAA (Music-Video Auto-Alignment), that automatically edits video to align with the rhythm of a given music track while preserving the original visual content. To enhance flexibility, we modularize the task into a two-step process in our MVAA: aligning motion keyframes with audio beats, followed by rhythm-aware video inpainting. Specifically, we first insert keyframes at timestamps aligned with musical beats, then use a frame-conditioned diffusion model to generate coherent intermediate frames, preserving the original video's semantic content. Since comprehensive test-time training can be time-consuming, we adopt a two-stage strategy: pretraining the inpainting module on a small video set to learn general motion priors, followed by rapid inference-time fine-tuning for video-specific adaptation. This hybrid approach enables adaptation within 10 minutes with one epoch on a single NVIDIA 4090 GPU using CogVideoX-5b-I2V as the backbone. Extensive experiments show that our approach can achieve high-quality beat alignment and visual smoothness.

cs.CV↗

arXiv2025

You Only Render Once: Enhancing Energy and Computation Efficiency of Mobile Virtual Reality

Mobile Virtual Reality (VR) is essential to achieving convenient and immersive human-computer interaction and realizing emerging applications such as Metaverse. However, existing VR technologies require two separate renderings of binocular images, causing a significant bottleneck for mobile devices with limited computing capability and power supply. This paper proposes an approach to rendering optimization for mobile VR called EffVR. By utilizing the per-pixel attribute, EffVR can generate binocular VR images from the monocular image through genuinely one rendering, saving half the computation over conventional approaches. Our evaluation indicates that, compared with the state-of-art, EffVRcan save 27% power consumption on average while achieving high binocular image quality (0.9679 SSIM and 34.09 PSNR) in mobile VR applications. Additionally, EffVR can increase the frame rate by 115.2%. These results corroborate EffVRsuperior computation/energy-saving performance, paving the road to a sustainable mobile VR. The source code, demo video, android app, and more are released anonymously at https://yoro-vr.github.io/

cs.GR↗

arXiv2025

A Transfer Learning Framework for Multilayer Networks via Model Averaging

Link prediction in multilayer networks is a key challenge in applications such as recommendation systems and protein-protein interaction prediction. While many techniques have been developed, most rely on assumptions about shared structures and require access to raw auxiliary data, limiting their practicality. To address these issues, we propose a novel transfer learning framework for multilayer networks using a bi-level model averaging method. A $K$-fold cross-validation criterion based on edges is used to automatically weight inter-layer and intra-layer candidate models. This enables the transfer of information from auxiliary layers while mitigating model uncertainty, even without prior knowledge of shared structures. Theoretically, we prove the optimality and weight convergence of our method under mild conditions. Computationally, our framework is efficient and privacy-preserving, as it avoids raw data sharing and supports parallel processing across multiple servers. Simulations show our method outperforms others in predictive accuracy and robustness. We further demonstrate its practical value through two real-world recommendation system applications.

stat.ML↗

arXiv2025

V2I-Calib++: A Multi-terminal Spatial Calibration Approach in Urban Intersections for Collaborative Perception

Urban intersections, dense with pedestrian and vehicular traffic and compounded by GPS signal obstructions from high-rise buildings, are among the most challenging areas in urban traffic systems. Traditional single-vehicle intelligence systems often perform poorly in such environments due to a lack of global traffic flow information and the ability to respond to unexpected events. Vehicle-to-Everything (V2X) technology, through real-time communication between vehicles (V2V) and vehicles to infrastructure (V2I), offers a robust solution. However, practical applications still face numerous challenges. Calibration among heterogeneous vehicle and infrastructure endpoints in multi-end LiDAR systems is crucial for ensuring the accuracy and consistency of perception system data. Most existing multi-end calibration methods rely on initial calibration values provided by positioning systems, but the instability of GPS signals due to high buildings in urban canyons poses severe challenges to these methods. To address this issue, this paper proposes a novel multi-end LiDAR system calibration method that does not require positioning priors to determine initial external parameters and meets real-time requirements. Our method introduces an innovative multi-end perception object association technique, utilizing a new Overall Distance metric (oDist) to measure the spatial association between perception objects, and effectively combines global consistency search algorithms with optimal transport theory. By this means, we can extract co-observed targets from object association results for further external parameter computation and optimization. Extensive comparative and ablation experiments conducted on the simulated dataset V2X-Sim and the real dataset DAIR-V2X confirm the effectiveness and efficiency of our method. The code for this method can be accessed at: https://github.com/MassimoQu/v2i-calib.

cs.RO↗

arXiv2025

VProChart: Answering Chart Question through Visual Perception Alignment Agent and Programmatic Solution Reasoning

Charts are widely used for data visualization across various fields, including education, research, and business. Chart Question Answering (CQA) is an emerging task focused on the automatic interpretation and reasoning of data presented in charts. However, chart images are inherently difficult to interpret, and chart-related questions often involve complex logical and numerical reasoning, which hinders the performance of existing models. This paper introduces VProChart, a novel framework designed to address these challenges in CQA by integrating a lightweight Visual Perception Alignment Agent (VPAgent) and a Programmatic Solution Reasoning approach. VPAgent aligns and models chart elements based on principles of human visual perception, enhancing the understanding of chart context. The Programmatic Solution Reasoning approach leverages large language models (LLMs) to transform natural language reasoning questions into structured solution programs, facilitating precise numerical and logical reasoning. Extensive experiments on benchmark datasets such as ChartQA and PlotQA demonstrate that VProChart significantly outperforms existing methods, highlighting its capability in understanding and reasoning with charts.

cs.CV↗

arXiv2025

Beyond the Surface: Measuring Self-Preference in LLM Judgments

Recent studies show that large language models (LLMs) exhibit self-preference bias when serving as judges, meaning they tend to favor their own responses over those generated by other models. Existing methods typically measure this bias by calculating the difference between the scores a judge model assigns to its own responses and those it assigns to responses from other models. However, this approach conflates self-preference bias with response quality, as higher-quality responses from the judge model may also lead to positive score differences, even in the absence of bias. To address this issue, we introduce gold judgments as proxies for the actual quality of responses and propose the DBG score, which measures self-preference bias as the difference between the scores assigned by the judge model to its own responses and the corresponding gold judgments. Since gold judgments reflect true response quality, the DBG score mitigates the confounding effect of response quality on bias measurement. Using the DBG score, we conduct comprehensive experiments to assess self-preference bias across LLMs of varying versions, sizes, and reasoning abilities. Additionally, we investigate two factors that influence and help alleviate self-preference bias: response text style and the post-training data of judge models. Finally, we explore potential underlying mechanisms of self-preference bias from an attention-based perspective. Our code and data are available at https://github.com/zhiyuanc2001/self-preference.

cs.CL↗

arXiv2025

RACE-Align: Retrieval-Augmented and Chain-of-Thought Enhanced Preference Alignment for Large Language Models

Large Language Models (LLMs) struggle with accuracy, domain-specific reasoning, and interpretability in vertical domains. Traditional preference alignment methods like Reinforcement Learning from Human Feedback (RLHF) and Direct Preference Optimization (DPO) often overlook the underlying knowledge sources and reasoning logic. This paper introduces RACE-Align (Retrieval-Augmented and Chain-of-Thought Enhanced Alignment), a novel framework designed to address these limitations. RACE-Align systematically constructs a binary preference dataset incorporating external knowledge support and explicit Chain-of-Thought (CoT) reasoning, then aligns LLMs using the DPO algorithm. The core innovation lies in its preference data construction strategy: it integrates AI-driven retrieval for factual grounding, enhancing knowledgeability and accuracy, and emphasizes the optimization of domain-specific CoT, treating the reasoning process itself as a key preference dimension. A multi-stage, AI-driven refinement pipeline cost-effectively generates these preference pairs. Experimental validation in Traditional Chinese Medicine (TCM) using Qwen3-1.7B as the base model demonstrates that RACE-Align significantly outperforms the original base model and a model fine-tuned only with Supervised Fine-Tuning (SFT). Improvements were observed across multiple dimensions, including answer accuracy, information richness, application of TCM thinking patterns, logicality and depth of reasoning, and interpretability. These findings suggest RACE-Align offers an effective pathway to enhance LLMs' knowledge application, reasoning reliability, and process transparency in complex vertical domains.

cs.CL↗

arXiv2025

Experimental Evidence That AI-Managed Workers Tolerate Lower Pay Without Demotivation

Experimental evidence on worker responses to AI management remains mixed, partly due to limitations in experimental fidelity. We address these limitations with a customized workplace in the Minecraft platform, enabling high-resolution behavioral tracking of autonomous task execution, and ensuring that participants approach the task with well-formed expectations about their own competence. Workers (N = 382) completed repeated production tasks under either human, AI, or hybrid management. An AI manager trained on human-defined evaluation principles systematically assigned lower performance ratings and reduced wages by 40\%, without adverse effects on worker motivation and sense of fairness. These effects were driven by a muted emotional response to AI evaluation, compared to evaluation by a human. The very features that make AI appear impartial may also facilitate silent exploitation, by suppressing the social reactions that normally constrain extractive practices in human-managed work.

cs.CY↗

arXiv2025

CogReact: A Reinforced Framework to Model Human Cognitive Reaction Modulated by Dynamic Intervention

Using deep neural networks as computational models to simulate cognitive process can provide key insights into human behavioral dynamics. Challenges arise when environments are highly dynamic, obscuring stimulus-behavior relationships. However, the majority of current research focuses on simulating human cognitive behaviors under ideal conditions, neglecting the influence of environmental disturbances. We propose CogReact, integrating drift-diffusion with deep reinforcement learning to simulate granular effects of dynamic environmental stimuli on human cognitive process. Quantitatively, it improves cognition modelling by considering temporal effect of environmental stimuli on cognitive process and captures both subject-specific and stimuli-specific behavioural differences. Qualitatively, it captures general trends in human cognitive process under stimuli, better than baselines. Our approach is examined in diverse environmental influences on various cognitive tasks. Overall, it demonstrates a powerful, data-driven methodology to simulate, align with, and understand the vagaries of human cognitive response in dynamic contexts.

cs.AI↗

arXiv2025

PhysReason: A Comprehensive Benchmark towards Physics-Based Reasoning

Large language models demonstrate remarkable capabilities across various domains, especially mathematics and logic reasoning. However, current evaluations overlook physics-based reasoning - a complex task requiring physics theorems and constraints. We present PhysReason, a 1,200-problem benchmark comprising knowledge-based (25%) and reasoning-based (75%) problems, where the latter are divided into three difficulty levels (easy, medium, hard). Notably, problems require an average of 8.1 solution steps, with hard requiring 15.6, reflecting the complexity of physics-based reasoning. We propose the Physics Solution Auto Scoring Framework, incorporating efficient answer-level and comprehensive step-level evaluations. Top-performing models like Deepseek-R1, Gemini-2.0-Flash-Thinking, and o3-mini-high achieve less than 60% on answer-level evaluation, with performance dropping from knowledge questions (75.11%) to hard problems (31.95%). Through step-level evaluation, we identified four key bottlenecks: Physics Theorem Application, Physics Process Understanding, Calculation, and Physics Condition Analysis. These findings position PhysReason as a novel and comprehensive benchmark for evaluating physics-based reasoning capabilities in large language models. Our code and data will be published at https:/dxzxy12138.github.io/PhysReason.

cs.AI↗

arXiv2025

S2R-Bench: A Sim-to-Real Evaluation Benchmark for Autonomous Driving

Safety is a long-standing and the final pursuit in the development of autonomous driving systems, with a significant portion of safety challenge arising from perception. How to effectively evaluate the safety as well as the reliability of perception algorithms is becoming an emerging issue. Despite its critical importance, existing perception methods exhibit a limitation in their robustness, primarily due to the use of benchmarks are entierly simulated, which fail to align predicted results with actual outcomes, particularly under extreme weather conditions and sensor anomalies that are prevalent in real-world scenarios. To fill this gap, in this study, we propose a Sim-to-Real Evaluation Benchmark for Autonomous Driving (S2R-Bench). We collect diverse sensor anomaly data under various road conditions to evaluate the robustness of autonomous driving perception methods in a comprehensive and realistic manner. This is the first corruption robustness benchmark based on real-world scenarios, encompassing various road conditions, weather conditions, lighting intensities, and time periods. By comparing real-world data with simulated data, we demonstrate the reliability and practical significance of the collected data for real-world applications. We hope that this dataset will advance future research and contribute to the development of more robust perception models for autonomous driving. This dataset is released on https://github.com/adept-thu/S2R-Bench.

cs.RO↗

arXiv2025

Qwen3 Technical Report

In this work, we present Qwen3, the latest version of the Qwen model family. Qwen3 comprises a series of large language models (LLMs) designed to advance performance, efficiency, and multilingual capabilities. The Qwen3 series includes models of both dense and Mixture-of-Expert (MoE) architectures, with parameter scales ranging from 0.6 to 235 billion. A key innovation in Qwen3 is the integration of thinking mode (for complex, multi-step reasoning) and non-thinking mode (for rapid, context-driven responses) into a unified framework. This eliminates the need to switch between different models--such as chat-optimized models (e.g., GPT-4o) and dedicated reasoning models (e.g., QwQ-32B)--and enables dynamic mode switching based on user queries or chat templates. Meanwhile, Qwen3 introduces a thinking budget mechanism, allowing users to allocate computational resources adaptively during inference, thereby balancing latency and performance based on task complexity. Moreover, by leveraging the knowledge from the flagship models, we significantly reduce the computational resources required to build smaller-scale models, while ensuring their highly competitive performance. Empirical evaluations demonstrate that Qwen3 achieves state-of-the-art results across diverse benchmarks, including tasks in code generation, mathematical reasoning, agent tasks, etc., competitive against larger MoE models and proprietary models. Compared to its predecessor Qwen2.5, Qwen3 expands multilingual support from 29 to 119 languages and dialects, enhancing global accessibility through improved cross-lingual understanding and generation capabilities. To facilitate reproducibility and community-driven research and development, all Qwen3 models are publicly accessible under Apache 2.0.

cs.CL↗

arXiv2025

SAR-GTR: Attributed Scattering Information Guided SAR Graph Transformer Recognition Algorithm

Utilizing electromagnetic scattering information for SAR data interpretation is currently a prominent research focus in the SAR interpretation domain. Graph Neural Networks (GNNs) can effectively integrate domain-specific physical knowledge and human prior knowledge, thereby alleviating challenges such as limited sample availability and poor generalization in SAR interpretation. In this study, we thoroughly investigate the electromagnetic inverse scattering information of single-channel SAR and re-examine the limitations of applying GNNs to SAR interpretation. We propose the SAR Graph Transformer Recognition Algorithm (SAR-GTR). SAR-GTR carefully considers the attributes and characteristics of different electromagnetic scattering parameters by distinguishing the mapping methods for discrete and continuous parameters, thereby avoiding information confusion and loss. Furthermore, the GTR combines GNNs with the Transformer mechanism and introduces an edge information enhancement channel to facilitate interactive learning of node and edge features, enabling the capture of robust and global structural characteristics of targets. Additionally, the GTR constructs a hierarchical topology-aware system through global node encoding and edge position encoding, fully exploiting the hierarchical structural information of targets. Finally, the effectiveness of the algorithm is validated using the ATRNet-STAR large-scale vehicle dataset.

eess.IV↗

arXiv2025

Mogao: An Omni Foundation Model for Interleaved Multi-Modal Generation

Recent progress in unified models for image understanding and generation has been impressive, yet most approaches remain limited to single-modal generation conditioned on multiple modalities. In this paper, we present Mogao, a unified framework that advances this paradigm by enabling interleaved multi-modal generation through a causal approach. Mogao integrates a set of key technical improvements in architecture design, including a deep-fusion design, dual vision encoders, interleaved rotary position embeddings, and multi-modal classifier-free guidance, which allow it to harness the strengths of both autoregressive models for text generation and diffusion models for high-quality image synthesis. These practical improvements also make Mogao particularly effective to process interleaved sequences of text and images arbitrarily. To further unlock the potential of unified models, we introduce an efficient training strategy on a large-scale, in-house dataset specifically curated for joint text and image generation. Extensive experiments show that Mogao not only achieves state-of-the-art performance in multi-modal understanding and text-to-image generation, but also excels in producing high-quality, coherent interleaved outputs. Its emergent capabilities in zero-shot image editing and compositional generation highlight Mogao as a practical omni-modal foundation model, paving the way for future development and scaling the unified multi-modal systems.

cs.CV↗

arXiv2025

Pigeonhole Stochastic Gradient Langevin Dynamics for Large Crossed Mixed Effects Models

Large crossed mixed effects models with imbalanced structures and missing data pose major computational challenges for standard Bayesian posterior sampling algorithms, as the computational complexity is usually superlinear in the number of observations. We propose two efficient subset-based stochastic gradient MCMC algorithms for such crossed mixed effects models, which facilitate scalable inference on both the variance components and regression coefficients. The first algorithm is developed for balanced design without missing observations, where we leverage the closed-form expression of the precision matrix for the full data matrix. The second algorithm, which we call the pigeonhole stochastic gradient Langevin dynamics (PSGLD), is developed for both balanced and unbalanced designs with potentially a large proportion of missing observations. Our PSGLD algorithm imputes the latent crossed random effects by running short Markov chains and then samples the model parameters of variance components and regression coefficients at each MCMC iteration. We provide theoretical guarantees by showing the convergence of the output distribution from the proposed algorithms to the target non-log-concave posterior distribution. A variety of numerical experiments based on both synthetic and real data demonstrate that the proposed algorithms can significantly reduce the computational cost of the standard MCMC algorithms and better balance the approximation accuracy and computational efficiency.

stat.CO↗