arXiv subjects

Xu Yang

Publications and source records attributed to Xu Yang.

At least 163 records · Page 9Linked to original sources

arXiv2026

Beyond Global Alignment: Fine-Grained Motion-Language Retrieval via Pyramidal Shapley-Taylor Learning

As a foundational task in human-centric cross-modal intelligence, motion-language retrieval aims to bridge the semantic gap between natural language and human motion, enabling intuitive motion analysis, yet existing approaches predominantly focus on aligning entire motion sequences with global textual representations. This global-centric paradigm overlooks fine-grained interactions between local motion segments and individual body joints and text tokens, inevitably leading to suboptimal retrieval performance. To address this limitation, we draw inspiration from the pyramidal process of human motion perception (from joint dynamics to segment coherence, and finally to holistic comprehension) and propose a novel Pyramidal Shapley-Taylor (PST) learning framework for fine-grained motion-language retrieval. Specifically, the framework decomposes human motion into temporal segments and spatial body joints, and learns cross-modal correspondences through progressive joint-wise and segment-wise alignment in a pyramidal fashion, effectively capturing both local semantic details and hierarchical structural relationships. Extensive experiments on multiple public benchmark datasets demonstrate that our approach significantly outperforms state-of-the-art methods, achieving precise alignment between motion segments and body joints and their corresponding text tokens. The code of this work will be released upon acceptance.

cs.CV↗

arXiv2026

SpecFLASH: A Latent-Guided Semi-autoregressive Speculative Decoding Framework for Efficient Multimodal Generation

Large language models and large multimodal models (LLMs and LMMs) deliver strong generative performance but suffer from slow decoding, a problem that becomes more severe when handling visual inputs, whose sequences typically contain many more tokens with lower information density than text. Speculative decoding accelerates LLM inference by letting a compact draft model propose candidate tokens that are selectively accepted by a larger target model, achieving speed-up without degrading quality. However, existing multimodal speculative decoding approaches largely ignore the structural characteristics of visual representations and usually rely on text-only draft models. In this paper, we introduce SpecFLASH, a speculative decoding framework tailored to LMMs that explicitly exploits multimodal structure when designing the draft model. We first mitigate redundancy in visual token sequences with a lightweight, latent-guided token compression module that compacts visual features while preserving semantics, and then leverage the co-occurrence and local correlations of visual entities via a semi-autoregressive decoding scheme that predicts multiple tokens in a single forward pass. Extensive experiments demonstrate that SpecFLASH consistently surpasses prior speculative decoding baselines, achieving up to $2.68\times$ speed-up on video captioning and $2.55\times$ on visual instruction tuning, relative to the original LMM. Our code is available here: https://github.com/ZihuaEvan/FlashSD/.

cs.CV↗

arXiv2026

Federated k-Means over Networks

We study federated clustering, where interconnected devices collaboratively cluster the data points of private local datasets. Focusing on hard clustering via the k-means principle, we formulate federated k-means as an instance of generalized total variation minimization (GTVMin). This leads to a federated k-means algorithm in which each device updates its local cluster centroids by solving a regularized k-means problem with a regularizer that enforces consistency between neighbouring devices. The resulting algorithm is privacy-friendly, as only aggregated information is exchanged.

cs.LG↗

arXiv2026

GeoSSA: Geometric Sparrow Search Algorithm for UAV Path Planning and Engineering Design Optimization

The Sparrow Search Algorithm (SSA), characterized by its simple structure and ease of implementation, nevertheless suffers from an insufficient balance between exploration and exploitation, making it prone to premature convergence and slow optimization progress. To address these shortcomings, this paper proposes a Geometric Sparrow Search Algorithm (GeoSSA). By integrating Good Nodes Set initialization, a Sine-Cosine Enhanced Producer position update strategy, and a Triangular-Walk Enhanced Edge Sparrow update strategy, GeoSSA significantly improves the global exploration ability, local exploitation efficiency, and convergence stability of the original SSA. To thoroughly validate the effectiveness of GeoSSA, we conducted ablation studies, qualitative analysis, and comparative experiments on 23 benchmark functions against state-of-the-art algorithms. Experimental results show that GeoSSA achieves the best or near-best performance in terms of average fitness, standard deviation, Wilcoxon tests, and Friedman rankings, with an Overall Effectiveness ($OE$) of 95.65\%. Its overall performance is significantly superior to all compared algorithms. In three-dimensional UAV path planning tasks, GeoSSA demonstrates excellent stability and superior path quality. In four categories of engineering design optimization problems, GeoSSA consistently attains the highest solution accuracy and strongest stability. GeoSSA not only exhibits outstanding global optimization performance on standard benchmark functions but also shows strong robustness and generalization ability in practical applications such as UAV path planning and engineering design. Therefore, GeoSSA provides an efficient and reliable solution framework for complex optimization problems.

cs.CE↗

arXiv2026

Antarctic Infrared Binocular Telescope: Early Data Release of observations in the 1.4 μm water-vapor-absorption band

Ground-based observations around 1.4 $μ$m are normally limited by strong absorption of telluric water-vapor. However, Dome A, Antarctica has exceptionally dry conditions that offer a unique opportunity for observations in this band. We designed a new filter covering 1.34--1.48 $μ$m, namely $W'$, and installed it on the Antarctic Infrared Binocular Telescope (AIRBT) at Dome A in 2025. AIRBT comprises two identical 15 cm optical tube assemblies and two InGaAs cameras equipped with $J$ and $W'$ filters, respectively. With this Early Data Release (EDR), we aim to evaluate the performance of the $W'$ band at Dome A to observe objects with water-vapor features. This EDR covers $\thicksim 20 \ \mathrm{deg^2}$ in the Galactic plane using $\thicksim 20,000$ images in three nights. For 2 s exposures, the 5 $σ$ limiting magnitude histogram peaks at $J \thicksim 11.5$ mag (Vega) and $W' \thicksim 9.9$ mag, respectively. The $J-W'$ vs $J-H$ color-color diagram distinguishes ultracool candidates with water-vapor-absorption features from reddened early type stars. Furthermore, later-type stars tend to exhibit stronger water-vapor absorption. Some sources show larger $ΔW'$ than $ΔJ$ across the three nights, which we attribute to variations of their water-vapor-absorption depth. We conclude that it will be efficient to search for ultracool stars and estimate their spectral subtypes using $W'$ band imaging at Dome A, where the atmospheric transmission is high and stable.

astro-ph.IM↗

arXiv2026

Towards Understanding Feature Learning in Parameter Transfer

Parameter transfer is a central paradigm in transfer learning, enabling knowledge reuse across tasks and domains by sharing model parameters between upstream and downstream models. However, when only a subset of parameters from the upstream model is transferred to the downstream model, there remains a lack of theoretical understanding of the conditions under which such partial parameter reuse is beneficial and of the factors that govern its effectiveness. To address this gap, we analyze a setting in which both the upstream and downstream models are ReLU convolutional neural networks (CNNs). Within this theoretical framework, we characterize how the inherited parameters act as carriers of universal knowledge and identify key factors that amplify their beneficial impact on the target task. Furthermore, our analysis provides insight into why, in certain cases, transferring parameters can lead to lower test accuracy on the target task than training a new model from scratch. To our best knowledge, our theory is the first to provide a dynamic analysis for parameter transfer and also the first to prove the existence of negative transfer theoretically. Numerical experiments and real-world data experiments are conducted to empirically validate our theoretical findings.

cs.LG↗

arXiv2026

Loupe: A Generalizable and Adaptive Framework for Image Forgery Detection

The proliferation of generative models has raised serious concerns about visual content forgery. Existing deepfake detection methods primarily target either image-level classification or pixel-wise localization. While some achieve high accuracy, they often suffer from limited generalization across manipulation types or rely on complex architectures. In this paper, we propose Loupe, a lightweight yet effective framework for joint deepfake detection and localization. Loupe integrates a patch-aware classifier and a segmentation module with conditional queries, allowing simultaneous global authenticity classification and fine-grained mask prediction. To enhance robustness against distribution shifts of test set, Loupe introduces a pseudo-label-guided test-time adaptation mechanism by leveraging patch-level predictions to supervise the segmentation head. Extensive experiments on the DDL dataset demonstrate that Loupe achieves state-of-the-art performance, securing the first place in the IJCAI 2025 Deepfake Detection and Localization Challenge with an overall score of 0.846. Our results validate the effectiveness of the proposed patch-level fusion and conditional query design in improving both classification accuracy and spatial localization under diverse forgery patterns. The code is available at https://github.com/Kamichanw/Loupe.

cs.CV↗

arXiv2026

Rotation Control Unlearning: Quantifying and Controlling Continuous Unlearning for LLM with The Cognitive Rotation Space

As Large Language Models (LLMs) become increasingly prevalent, their security vulnerabilities have already drawn attention. Machine unlearning is introduced to seek to mitigate these risks by removing the influence of undesirable data. However, existing methods not only rely on the retained dataset to preserve model utility, but also suffer from cumulative catastrophic utility loss under continuous unlearning requests. To solve this dilemma, we propose a novel method, called Rotation Control Unlearning (RCU), which leverages the rotational salience weight of RCU to quantify and control the unlearning degree in the continuous unlearning process. The skew symmetric loss is designed to construct the existence of the cognitive rotation space, where the changes of rotational angle can simulate the continuous unlearning process. Furthermore, we design an orthogonal rotation axes regularization to enforce mutually perpendicular rotation directions for continuous unlearning requests, effectively minimizing interference and addressing cumulative catastrophic utility loss. Experiments on multiple datasets confirm that our method without retained dataset achieves SOTA performance.

cs.LG↗

arXiv2026

Convergence Analysis of PINNs for Fractional Diffusion Equations in Bounded Domains

We establish the convergence of physics-informed neural networks (PINNs) for time-dependent fractional diffusion equations posed on bounded domains. The presence of fractional Laplacian operators introduces nonlocal behavior and regularity constraints, and standard neural network approximations do not naturally enforce the associated spectral boundary conditions. To address this challenge, we introduce a spectrally-defined mollification strategy that preserves the structure of the nonlocal operator while ensuring boundary compatibility. This enables the derivation of rigorous energy estimates in Sobolev spaces. Our results rely on analytical tools from PDE theory, highlighting the compatibility of PINN approximations with classical energy estimates for nonlocal equations. We prove convergence of the PINN approximation in any space-time Sobolev norm $H^k$ (with $k \in \N)$. The analysis highlights the role of mollified residuals in enabling theoretical guarantees for neural-network-based solvers of nonlocal PDEs.

math.NA↗

arXiv2025

Effective Online Exam Proctoring by Combining Lightweight Face Detection and Deep Recognition

Online exams conducted via video conferencing platforms such as Zoom have become widespread, yet ensuring exam integrity remains challenging due to the difficulty of monitoring multiple video feeds in real time. We present iExam, an online exam proctoring and analysis system that combines lightweight real-time face detection with deep face recognition for postexam analysis. iExam assists invigilators by monitoring student presence during exams and identifies abnormal behaviors, such as face disappearance, face rotation, and identity substitution, from recorded videos. The system addresses three key challenges: (i)efficient real-time video capture and analysis, (ii) automated student identity labeling using enhanced OCR on dynamic Zoom name tags, and (iii) resource-efficient training and inference on standard teacher devices. Extensive experiments show that iExam achieves 90.4% accuracy in real-time face detection and 98.4% accuracy in post-exam recognition with low overhead, demonstrating its practicality and effectiveness for online exam proctoring.

cs.CV↗

arXiv2025

Energy-Efficient Omnidirectional Locomotion for Wheeled Quadrupeds via Predictive Energy-Aware Nominal Gait Selection

Wheeled-legged robots combine the efficiency of wheels with the versatility of legs, but face significant energy optimization challenges when navigating diverse environments. In this work, we present a hierarchical control framework that integrates predictive power modeling with residual reinforcement learning to optimize omnidirectional locomotion efficiency for wheeled quadrupedal robots. Our approach employs a novel power prediction network that forecasts energy consumption across different gait patterns over a 1-second horizon, enabling intelligent selection of the most energy-efficient nominal gait. A reinforcement learning policy then generates residual adjustments to this nominal gait, fine-tuning the robot's actions to balance energy efficiency with performance objectives. Comparative analysis shows our method reduces energy consumption by up to 35\% compared to fixed-gait approaches while maintaining comparable velocity tracking performance. We validate our framework through extensive simulations and real-world experiments on a modified Unitree Go1 platform, demonstrating robust performance even under external disturbances. Videos and implementation details are available at \href{https://sites.google.com/view/switching-wpg}{https://sites.google.com/view/switching-wpg}.

cs.RO↗

arXiv2025

Revealing Perception and Generation Dynamics in LVLMs: Mitigating Hallucinations via Validated Dominance Correction

Large Vision-Language Models (LVLMs) have shown remarkable capabilities, yet hallucinations remain a persistent challenge. This work presents a systematic analysis of the internal evolution of visual perception and token generation in LVLMs, revealing two key patterns. First, perception follows a three-stage GATE process: early layers perform a Global scan, intermediate layers Approach and Tighten on core content, and later layers Explore supplementary regions. Second, generation exhibits an SAD (Subdominant Accumulation to Dominant) pattern, where hallucinated tokens arise from the repeated accumulation of subdominant tokens lacking support from attention (visual perception) or feed-forward network (internal knowledge). Guided by these findings, we devise the VDC (Validated Dominance Correction) strategy, which detects unsupported tokens and replaces them with validated dominant ones to improve output reliability. Extensive experiments across multiple models and benchmarks confirm that VDC substantially mitigates hallucinations.

cs.CV↗

arXiv2025

A geometric approach to predicting plasticity in disordered solids

It was recently shown that vortex-like topological defects with negative winding number in the vibrational modes of a two-dimensional glass under quasistatic shear correlate strongly with plastic events, offering a promising route to predict them. However, many of these vortices, a number that actually grows quadratically with mode frequency, are entirely unrelated to plasticity and arise simply from the underlying plane-wave structure of the modes. This raises doubts about the fundamental relevance of such defects to plastic rearrangements and limits their predictive power. Here, we introduce a geometrical filter based on the Nye dislocation density that, when applied to the vibrational modes, removes these spurious defects and reveals the true plastic precursors. Using simulations of a two-dimensional model glass, we show that this filtered approach consistently outperforms the conventional vortex-based method, particularly at small strains and when focusing on genuine plastic stress drops, offering a more robust tool to predicting plasticity in glasses from their undeformed initial state.

cond-mat.soft↗

arXiv2025

Ultrahigh-Q chiral resonances empowered by multi-head attention deep learning

High quality (Q) factor optical chiral resonators are indispensable for many chiral photonic devices. Designing ultrahigh Q-factors in chiral metasurfaces traditionally relies on extensive parameter scanning, which is time-consuming and inefficient. While deep learning now provides a rapid design alternative, conventional models still face challenges in accurately predicting ultrahigh Q-factor spectral characteristics. In this study, we introduce a multi-head attention network (MuHAN) to accelerate the design of ultrahigh Q-factor optical chiral resonators in bilayer metasurfaces. MuHAN achieves forward spectral predictions in approximately 10ms, thousands of times faster than finite-difference time-domain simulations, boasting 99.85% and 99.9% accuracy for forward and inverse predictions, respectively. By transferring the learned physical principles, we perform inverse design of nanoscale structures with ultrahigh Q-factors (up to 2.9910E5) based on chiral quasi-bound states in the continuum (quasi-BICs) at minimal computational cost. Our rapid design tool, based on MuHAN, enables high-performance encryption imaging, bridging deep learning with high-Q chiral metasurfaces for advanced sensing, laser, and detection applications.

physics.optics↗

arXiv2025

NAWOA-XGBoost: A Novel Model for Early Prediction of Academic Potential in Computer Science Students

Whale Optimization Algorithm (WOA) suffers from limited global search ability, slow convergence, and tendency to fall into local optima, restricting its effectiveness in hyperparameter optimization for machine learning models. To address these issues, this study proposes a Nonlinear Adaptive Whale Optimization Algorithm (NAWOA), which integrates strategies such as Good Nodes Set initialization, Leader-Followers Foraging, Dynamic Encircling Prey, Triangular Hunting, and a nonlinear convergence factor to enhance exploration, exploitation, and convergence stability. Experiments on 23 benchmark functions demonstrate NAWOA's superior optimization capability and robustness. Based on this optimizer, an NAWOA-XGBoost model was developed to predict academic potential using data from 495 Computer Science undergraduates at Macao Polytechnic University (2009-2019). Results show that NAWOA-XGBoost outperforms traditional XGBoost and WOA-XGBoost across key metrics, including Accuracy (0.8148), Macro F1 (0.8101), AUC (0.8932), and G-Mean (0.8172), demonstrating strong adaptability on multi-class imbalanced datasets.

cs.CE↗

arXiv2025

ERF-BA-TFD+: A Multimodal Model for Audio-Visual Deepfake Detection

Deepfake detection is a critical task in identifying manipulated multimedia content. In real-world scenarios, deepfake content can manifest across multiple modalities, including audio and video. To address this challenge, we present ERF-BA-TFD+, a novel multimodal deepfake detection model that combines enhanced receptive field (ERF) and audio-visual fusion. Our model processes both audio and video features simultaneously, leveraging their complementary information to improve detection accuracy and robustness. The key innovation of ERF-BA-TFD+ lies in its ability to model long-range dependencies within the audio-visual input, allowing it to better capture subtle discrepancies between real and fake content. In our experiments, we evaluate ERF-BA-TFD+ on the DDL-AV dataset, which consists of both segmented and full-length video clips. Unlike previous benchmarks, which focused primarily on isolated segments, the DDL-AV dataset allows us to assess the model's performance in a more comprehensive and realistic setting. Our method achieves state-of-the-art results on this dataset, outperforming existing techniques in terms of both accuracy and processing speed. The ERF-BA-TFD+ model demonstrated its effectiveness in the "Workshop on Deepfake Detection, Localization, and Interpretability," Track 2: Audio-Visual Detection and Localization (DDL-AV), and won first place in this competition.

cs.AI↗

arXiv2025

An Enhanced Whale Optimization Algorithm with Log-Normal Distribution for Optimizing Coverage of Wireless Sensor Networks

Wireless Sensor Networks (WSNs) are essential for monitoring and communication in complex environments, where coverage optimization directly affects performance and energy efficiency. However, traditional algorithms such as the Whale Optimization Algorithm (WOA) often suffer from limited exploration and premature convergence. To overcome these issues, this paper proposes an enhanced WOA which is called GLNWOA. GLNWOA integrates a log-normal distribution model into WOA to improve convergence dynamics and search diversity. GLNWOA employs a Good Nodes Set initialization for uniform population distribution, a Leader Cognitive Guidance Mechanism for efficient information sharing, and an Enhanced Spiral Updating Strategy to balance global exploration and local exploitation. Tests on benchmark functions verify its superior convergence accuracy and robustness. In WSN coverage optimization, deploying 25 nodes in a 60 m $\times$ 60 m area achieved a 99.0013\% coverage rate, outperforming AROA, WOA, HHO, ROA, and WOABAT by up to 15.5\%. These results demonstrate that GLNWOA offers fast convergence, high stability, and excellent optimization capability for intelligent network deployment.

cs.CE↗

arXiv2025

GraphIC: A Graph-Based In-Context Example Retrieval Model for Multi-Step Reasoning

In-context learning (ICL) enhances large language models (LLMs) by incorporating demonstration examples, yet its effectiveness heavily depends on the quality of selected examples. Current methods typically use text embeddings to measure semantic similarity, which often introduces bias in multi-step reasoning tasks. This occurs because text embeddings contain irrelevant semantic information and lack deeper reasoning structures. To address this, we propose GraphIC, a graph-based retrieval model that leverages reasoning-aware representation and specialized similarity metric for in-context example retrieval. GraphIC first constructs thought graphs-directed, node-attributed graphs that explicitly model reasoning steps and their dependencies-for candidate examples and queries. This approach filters out superficial semantics while preserving essential reasoning processes. Next, GraphIC retrieves examples using a novel similarity metric tailored for these graphs, capturing sequential reasoning patterns and asymmetry between examples. Comprehensive evaluations across mathematical reasoning, code generation, and logical reasoning tasks demonstrate that GraphIC outperforms 10 baseline methods. Our results highlight the importance of reasoning-aware retrieval in ICL, offering a robust solution for enhancing LLM performance in multi-step reasoning scenarios.

cs.AI↗