Searcharxiv⌕ Search

arXiv subjects

Yumeng Zhang

Publications and source records attributed to Yumeng Zhang.

At least 19 recordsLinked to original sources

Morphology-Decoupled Segmentation for Multiple Tunnel Defects Under Complex Environments

Automated tunnel-lining inspection must segment different defect geometries: cracks are thin, elongated, and connectivity-sensitive, whereas leakage, lining falling off, and segment damage are spatially extensive and context-dependent. Existing approaches generally apply a common decoder across defect classes or introduce boundary, topology, and directional mechanisms as isolated enhancements. This paper introduces Morphology-Decoupled Segmentation (MDS), in which defect morphology becomes the organising principle for feature routing, decoding, and interaction. A supervised routing module partitions multiscale encoder features into complementary linear and areal streams. The linear stream employs Deformable Strip Attention with image-conditioned direction hypotheses to preserve crack continuity without assuming fixed horizontal or vertical orientations, while the areal stream combines Pyramid Pooling with feature-pyramid fusion to capture diffuse boundaries and regional context. The representations are coupled through gated cross-morphology attention, allowing complementary information exchange while retaining morphology-specific features. A resolution-adaptive tiling and probability-blending pipeline preserves fine defects across variable-resolution imagery. MDS achieves 81.17% mIoU on the internal validation set and improves foreground image-level recognition over a backbone-matched SegDINO model, primarily for areal defects. Controlled ablations demonstrate the complementary contributions of routing supervision and cross-stream interaction while revealing a trade-off between areal-class performance and crack specialisation. The submitted model achieved 67.55% mIoU on an organiser-controlled hidden test, and the team placed second. Selective dynamic INT8 quantisation reduced model storage by 2.83 times with negligible accuracy loss.

cs.CV↗

Learning Graph Beliefs for Active Radio Mapping With Decision-Certified Multi-Step Sensing

Radio maps provide essential wireless propagation knowledge for network optimization and autonomous platforms. Acquiring dense measurements, however, is prohibitively costly, requiring active sampling to prioritize informative measurements. Existing active sampling methods typically rely on variances from prescribed spatial models or learned pointwise uncertainty, which may not fully capture the spatial correlation of different measurements. This paper proposes GraphBARC, a graph-structured active radio map construction framework that integrates learned spatial modeling with analytical Bayesian inference. A graph network learns correlated node and edge factors using only acquired measurements, inducing a joint full-map belief that supports analytical observation updates without retraining. Rather than immediate one-step acquisition, it adopts a nonmyopic sensing strategy that evaluates joint map-wide Bayes risk reduction over feasible multi-step routes, accounting for information redundancy among planned measurements. We reuse candidate responses and derive residual-based bounds directly on route-value differences, retaining shared response-error structure to avoid uniformly high-accuracy evaluation. An error-aware continuation bound screens unpromising prefixes, after which only responses needed for unresolved comparisons are refined. We further establish a computable decision-gap bound and show that certified exclusions remain valid during refinement. Experiments show that GraphBARC reduces mean root-mean-square error (RMSE) by over 16% relative to the evaluated baselines and core planning time by 56% compared with the fixed-accuracy, unpruned implementation.

cs.NI↗

Enhancing Autoregressive Video Generation via Representation Adversarial Distillation

Few-step autoregressive video generation enables efficient streaming synthesis, but errors introduced in early temporal blocks are reused as context and can propagate through subsequent rollouts, leading to detail degradation, structural drift, and unstable motion. Existing distribution matching distillation (DMD) primarily aligns student and teacher distributions in diffusion latent space, but provides no direct supervision over the perceptual quality of decoded videos. We introduce Radian, a representation-space adversarial distillation framework that complements on-policy DMD with real-data adversarial supervision in the feature space defined by a frozen visual foundation model (VFM). During training, Radian sparsely decodes frames from autoregressive student rollouts, extracts multi-level visual representations, and applies lightweight discriminator heads to distinguish generated outputs from real video frames. The DMD objective anchors the student to the pretrained teacher, while the representation-space adversarial objective supplies complementary perceptual and semantic gradients that promote high-quality modes. These additional components are discarded after training, leaving the generator architecture and inference-time denoising budget unchanged. Experiments on Wan2.1-1.3B cover four-step chunk-wise, one-step frame-wise, and minute-long autoregressive generation. Our method achieves a VBench Total of 0.8444 and a VideoAlign Total of 0.8033 under four-step generation, and improves VBench-Long from 0.7805 to 0.8041 over Rolling Forcing while using fewer denoising steps. Controlled comparisons across image, video, and diffusion representations further indicate that the choice of representation spaces induces distinct adversarial signals, and external VFM gradients complement DMD more effectively than adversarial supervision derived from diffusion-internal features.

cs.CV↗

GPARA: Graph-Posterior-Aligned Refinement and Active Acquisition for Grounding Diffusion Priors

Active grounding of a frozen diffusion prior requires jointly determining where new measurements should be taken and how they should be used to refine the current reconstruction. Posterior-ensemble-based methods can estimate acquisition utility from generated samples, but require repeated ensemble generation as observations accumulate and capture posterior geometry only through empirical statistics. This paper proposes GPARA, which learns a context-dependent graph surrogate over diffusion prediction residuals, inducing an explicitly reusable posterior response operator that propagates measurement innovations to unobserved variables and evaluates candidate measurements through weighted posterior-risk reduction. Under the matched surrogate, we show that the same response operator also determines expected one-step acquisition benefit and yields an analytic ranking consistent with expected reconstruction improvement. A bounded learned residual calibrates the analytic utility to account for surrogate mismatch, while a small prior ensemble is generated once and reconditioned to update risk weights without repeated diffusion posterior sampling during acquisition. Experiments on two reconstruction tasks spanning physical field and computer vision show consistent improvements in refinement and active acquisition over the evaluated baselines. Ablations further support the complementary roles of step-wise graph refinement, adaptive risk weighting, and analytically anchored calibration.

cs.LG↗

TTRSD: Test-Time Reinforcement Learning with Self-Distillation for Vision-Language Models

Test-time reinforcement learning enables vision-language models (VLMs) to adapt using unlabeled inputs. However, repeated sampling under fixed visual conditions can reinforce shared perceptual errors, while sequence-level rewards fail to isolate visual perception the foundational bottleneck that anchors multimodal reasoning risking the degradation of pre-trained reasoning capabilities. We propose TTRSD, a test-time reinforcement learning framework combining multi-view answer-level self-distillation with visual contrastive token selection. A shared policy aggregates teacher predictions across original, cropped, and downsampled views into an answer distribution. Student trajectories generated from the original image receive rewards based on the support for their final answers in this distribution. To allocate this feedback precisely toward perceptual bottlenecks, we compare the log-probabilities of the same sampled tokens under original and visually ablated inputs while holding their textual prefixes fixed, selecting visually sensitive positions for policy-gradient updates. TTRSD separates update direction, determined by group-relative advantages, from update position, determined by visual sensitivity, without requiring ground-truth labels, external verifiers, or a separate teacher. With only 20 unlabeled adaptation samples, TTRSD improves performance across seven benchmarks and three VLMs, raising InternVL3-2B's MMMU accuracy from 35.79% to 49.32%(+13.53%), demonstrating cross-dataset generalization while preserving inherent reasoning integrity.

cs.CV↗

EMoG: Emotion-Modulated Gait Generation for Expressive Humanoid Locomotion

Existing humanoid locomotion systems primarily focus on stability and task execution, while integrating expressiveness with explicit locomotion control remains challenging. We propose EMoG, an emotion-modulated gait generation framework for expressive humanoid locomotion. EMoG introduces an emotional-style code with continuously adjustable intensity. Conditioned on this code and physical commands, a lightweight MLP generates expressive, command-consistent periodic gait trajectories in real time, which are tracked by a unified reinforcement learning policy for physical execution. To support training, we collect a large-scale emotion-annotated gait dataset from professional performers and develop an automated pipeline to extract physically consistent periodic gait cycles. EMoG also integrates an LLM-based parser that converts free-form language into emotional style and motion parameters for interactive control. Experiments demonstrate that our system achieves continuous gait-style modulation with perceptible expressive cues while maintaining command tracking. EMoG provides a practical approach to parameterized emotional-style walking for human-robot interaction.

cs.RO↗

Resilience in Trustworthy Wireless Systems

Resilience has emerged as a fundamental capability for future wireless systems operating in dynamic and uncertain environments. Although resilience has attracted growing attention across academia, industry, and standardization, its conceptual scope, enabling mechanisms, and realization techniques remain fragmented. This paper presents a systematic framework of resilience in wireless systems from a trustworthiness perspective. We first formalize the concept of resilience and distinguish it from related uncertainty-aware terminologies, including reliability, robustness, adaptability, survivability, and recoverability. We then establish a hierarchical framework that organizes resilience into capability dimensions and enabling mechanisms, and that quantifies resilience through different technical aspects. We further use physical links and unmanned aerial vehicle networks as representative wireless scenarios to demonstrate how resilience can be systematically realized through the joint design of architectures, operations, and algorithms. Finally, we examine the fundamental trade-offs in resilience engineering: Improving resilience generally incurs costs in resource efficiency, nominal performance, information acquisition, implementation complexity, and latency.

cs.IT↗

A functional joint model with baseline functional covariates: linking sitting accumulation patterns to physical function and mortality among older women

In large-scale epidemiological studies, it is often of interest to investigate joint relationships between longitudinal and time-to-event outcomes with exposures that are trajectories or functions. Our motivation study is the Objective Physical Activity and Cardiovascular Health (OPACH) Study, which collected accelerometry-measured physical activity in 6,489 older women. One of the scientific aims is to understand sedentary behavior accumulation patterns and its association with physical function (longitudinal) and mortality (time-to-event). We propose a novel approach that first converts raw accelerometry data into daily sitting bout accumulation profiles, which are treated as functional covariates, and then develop a functional joint model for longitudinal and time-to-event outcomes that incorporates a baseline functional covariate. The longitudinal process is modeled using functional data methods and linked to the survival process through functional principal component scores. Both sub-models include interpretable linear scalar-on-function regression coefficients to capture flexible dose-response associations between sitting accumulated across varying bout durations and health outcomes. Estimation is carried out via an efficient expectation-maximization (EM) algorithm with penalized spline approximations. Simulation studies demonstrate accurate parameter estimation and reliable model selection. Application to the OPACH data reveals flexible and interpretable dose-response relationships between sitting bout durations, physical function, and mortality.

stat.AP↗

Omni-LiveAvatar: Minute-Level Real-Time Streaming Joint Audio-Video Avatar Generation

Joint audio-video generative models serve as foundation for immersive and interactive digital-human generation. Nevertheless, most existing models rely on bidirectional attention and multi-step denoising and can generate only short clips, making them unsuitable for real-time interaction over extended durations. We present Omni-LiveAvatar, the first framework for minute-level, real-time streaming joint audio-video avatar generation. Specifically, we propose (1) a progressive autoregressive distillation pipeline that transfers a large bidirectional joint audio-video diffusion model into a few-step autoregressive generator without auxiliary stabilization mechanisms; (2) a synchronized audio-video long-short-term memory that preserves global consistency under a bounded memory budget; and (3) a hierarchical rolling prompt planning strategy that enables coherent semantic evolution and seamless prompt transitions. Extensive experiments show that Omni-LiveAvatar generates high-quality, synchronized minute-level avatars in real time. In terms of speed, it achieves a 33$\times$ generation speedup over its teacher, LTX-2, on a single NVIDIA H200 GPU; in terms of generation quality, it outperforms accelerated baselines across visual quality, audio quality, cross-modal synchronization, and human fidelity. Our code is available at https://github.com/Aoko955/Omni-LiveAvatar.

cs.MM↗

Flash-VAED: Plug-and-Play VAE Decoders for Efficient Video Generation

Latent diffusion models have enabled high-quality video synthesis, yet their inference remains costly and time-consuming. As diffusion transformers become increasingly efficient, the latency bottleneck inevitably shifts to VAE decoders. To reduce their latency while maintaining quality, we propose a universal acceleration framework for VAE decoders that preserves full alignment with the original latent distribution. Specifically, we propose (1) an independence-aware channel pruning method to effectively mitigate severe channel redundancy, and (2) a stage-wise dominant operator optimization strategy to address the high inference cost of the widely used causal 3D convolutions in VAE decoders. Based on these innovations, we construct a Flash-VAED family. Moreover, we design a three-phase dynamic distillation framework that efficiently transfers the capabilities of the original VAE decoder to Flash-VAED. Extensive experiments on Wan and LTX-Video VAE decoders demonstrate that our method outperforms baselines in both quality and speed, achieving approximately a 6$\times$ speedup while maintaining the reconstruction performance up to 96.9%. Notably, Flash-VAED accelerates the end-to-end generation pipeline by up to 36% with negligible quality drops on VBench-2.0. Our code is available at https://github.com/Aoko955/Flash-VAED.

cs.CV↗

MAFIA: Query-Only Memory Attacks via Probing and Factual Injection against Audited LLM Agents

Memory-augmented LLM agents rely on rich context for long-horizon reasoning and acting, yet their memory modules expose a persistent attack surface for malicious records, making the study of memory poisoning threats imperative. However, existing query-only attacks often fail to remain effective in two realistic and prevalent settings: large-scale benign memory pools and active input auditing. Consequently, current approaches fall short when facing the dual challenges of high retrieval competitiveness and rigorous semantic checks. To overcome these limitations, we propose MAFIA, a query-only Memory Attack framework via probing and Factual Injection against Audit, tailored to this extended threat model. Specifically, MAFIA introduces: (1) a placement strategy that ensures retrieval-competitive injection via memory probing, budget allocation, and scheduling; and (2) a payload design that bypasses audits using compact factual cloaks, preserving malicious effects while maintaining high semantic similarity. Extensive evaluations reveal that MAFIA achieves up to a 90.7% attack success rate while suppressing audit detection from a peak of 83.3% to at most 7.4%, exposing critical vulnerabilities across agentic memory systems. Code will be made publicly available at https://github.com/JiamingChen1234/MAFIA.

cs.AI↗

Generation Models Know Space: Unleashing Implicit 3D Priors for Scene Understanding

While Multimodal Large Language Models demonstrate impressive semantic capabilities, they often suffer from spatial blindness, struggling with fine-grained geometric reasoning and physical dynamics. Existing solutions typically rely on explicit 3D modalities or complex geometric scaffolding, which are limited by data scarcity and generalization challenges. In this work, we propose a paradigm shift by leveraging the implicit spatial prior within large-scale video generation models. We posit that to synthesize temporally coherent videos, these models inherently learn robust 3D structural priors and physical laws. We introduce VEGA-3D (Video Extracted Generative Awareness), a plug-and-play framework that repurposes a pre-trained video diffusion model as a Latent World Simulator. By extracting spatiotemporal features from intermediate noise levels and integrating them with semantic representations via a token-level adaptive gated fusion mechanism, we enrich MLLMs with dense geometric cues without explicit 3D supervision. Extensive experiments across 3D scene understanding, spatial reasoning, and embodied manipulation benchmarks demonstrate that our method outperforms state-of-the-art baselines, validating that generative priors provide a scalable foundation for physical-world understanding. Code is publicly available at https://github.com/H-EmbodVis/VEGA-3D.

cs.CV↗

Sparse Channel Estimation for Pixel Antennas: Addressing the Pilot Rank Deficiency

Composed of multiple interconnected pixels controlled by on/off RF switches, the pixel antenna can generate reconfigurable radiation patterns that can be further exploited to construct diverse pilot sequences for effective channel estimation. However, such pilot sequences inherently have rank deficiency, making it difficult to effectively and efficiently acquire the full channel state information (CSI) across all available radiation patterns. To tackle this difficulty, we consider a sparse environment with a limited number of propagation paths for a pixel antenna system, where a user equipped with a pixel antenna transmits only a limited number of pilots to recover the CSI under all radiation patterns. The proposed algorithm exploits the limited number of propagation paths that are invariant with the pixel antenna patterns, and then formulates the full channel estimation as a sparse recovery problem in the angular domain solved by Generalized Approximate Message Passing (GAMP). Moreover, to mitigate the rank deficiency of pilot sequences, we additionally incorporate a Multipath Matching Pursuit (MMP) algorithm for robust initialization. The overall proposed scheme, termed MMP-GAMP, achieves higher estimation accuracy than other algorithm baselines, while requiring lower pilot overhead.

eess.SP↗

Looped World Models

Current world models face a fundamental tension: faithful long-horizon simulation demands deep computation, but deeper models are expensive to deploy and prone to compounding errors. We resolve this by introducing Looped World Models (LoopWM), which are the first looped architectures for world modelling. Our method iteratively refines latent environment states through a parameter-shared transformer block. This yield up to 100x parameter efficiency over conventional approaches with adaptive computation that automatically scales depth to match the complexity of each prediction step. Orthogonal to scaling model size and training data, LoopWM establishes iterative latent depth as a new scaling axis for world simulation, which might significantly push the community forward.

cs.LG↗

See More, Think Deeper: Query-Expanded Visual Evidence and Answer-Clue Guided Reflection for Long Video Understanding

Recent advances in Video Large Language Models (Video-LLMs) have enabled performance on long-video understanding tasks. However, existing methods still face two key limitations: evidence acquisition often relies on a single search intent, and answer generation lacks an effective visual feedback mechanism. To address these limitations, we propose \textbf{CoVER}, a Comprehensive Visual Evidence and Reflection framework for long-video understanding. CoVER enables Video-LLMs to \textbf{See More} by dynamically gathering query-expanded visual evidence, and \textbf{Think Deeper} by verifying draft answers with effective answer-specific visual feedback. Together, these mechanisms shift long-video understanding from answer-centric generation to evidence-centric and visually verifiable reasoning. Experimental results show that CoVER-7B substantially outperforms models with the same parameter scale and even surpasses state-of-the-art closed-source models on certain metrics.

cs.CV↗

A Visual Analytics System for Interactive Exploration of Historical Painter Cohorts

Painter cohort analysis has long been regarded as a key lens for studying how painting artistic styles develop and transmit across generations. Through a two-year collaboration with art historians, we identify key challenges in traditional painter cohort research: the unstructured characteristic of painter features, the entangled complexity of inheritance relationships, and the cognitively demanding nature of cohort definition and validation. To solve these challenges, we propose HPC-Vis, a visual analytics system for interactive exploration of historical painter cohorts. An improved cohort analytical workflow is designed to integrate structured feature construction, visualization-assisted exploration, algorithm-based recommendation, and unified cohort management. Based on this workflow, we develop three core computational modules: a multi-scale artistic feature construction method that leverages LLMs to extract and organize hierarchical style features from unstructured historical texts, an inheritance reconstruction algorithm that transforms the entangled multi-parent inheritance network into a clear hierarchical forest structure, and a recommendation model that identifies core features of the cohort and recommends cohort members via painter relevance assessment. To support smooth interactive exploration, we further design a set of novel visualizations with multidimensional collaboration, especially an inheriting mountain view inspired by traditional Chinese landscape paintings, and a foldable doughnut chart for hierarchical artistic style labels. HPC-Vis is evaluated and validated through case studies, user studies, and technical evaluations, demonstrating its effectiveness in supporting painter cohort exploration and in providing visual insights for art historical research.

cs.HC↗

GeoGS-CE: Learning Delay--Beam Channel Priors with 3D Gaussians for High-Mobility Scenarios

Wideband channel estimation (CE) in high-mobility scenarios remains challenging because channel responses vary rapidly, while practical systems can allocate only sparse pilots to accommodate dense users. Fortunately, many high-mobility environments, such as high-speed railways, exhibit scheduled trajectories, predictable velocities, and a limited number of dominant propagation paths. These properties induce a delay--beam power spectrum that is more stable than the instantaneous complex channel frequency response (CFR), less sensitive to the random phase coherence, and rich in geometric information. To exploit such environmental properties, we propose GeoGS-CE, a two-stage channel estimation framework for sparse-pilot high-mobility scenarios. In the offline stage, GeoGS-CE jointly models: 1) a scene-level 3D Gaussian representation that captures the non-line-of-sight (NLoS) geometric scattering support, and 2) a leakage-aware differentiable wireless rendering process that maps the NLoS Gaussians, together with an explicit virtual line-of-sight (LoS) component, to the measured delay--beam power spectrum, while accounting for practical OFDM delay and array leakage effects. In the online stage, the delay--beam power spectrum is predicted for each user location and used as a strong covariance prior, enabling accurate full-band and full-array CFR reconstruction and tracking through a linear MMSE estimator. Simulations based on channels generated from a segment of the Guangshen high-speed railway show that the proposed geometric prior substantially improves CFR reconstruction over pilot-only and non-geometric baselines.

cs.IT↗

Distribution Corrected Offline Data Distillation for Large Language Models

Distilling reasoning traces from strong large language models into smaller ones is a promising route to improve intelligence in resource-constrained settings. Existing approaches face a fundamental trade-off: offline distillation from teacher-generated traces provides high-quality, sample-efficient supervision but suffers from distributional drift: during training, the student model conditions on teacher-generated prefixes, whereas during inference the student autoregresses on self-generated prefixes, leading to compounding errors over long reasoning trajectories. Meanwhile, on-policy or self-distillation methods better match the student's inference-time distribution, but require costly online sampling and often produce low-quality traces in early training. We propose a principled offline reasoning distillation framework that preserves the efficiency and supervision quality of offline teacher-generated data while correcting teacher-student distribution drift. It adaptively emphasizes teacher supervision that is better aligned with the student's on-policy distribution. Evaluations on mathematical reasoning benchmarks of GSM8K, MATH, MATH500, and harder held-out competition-style tasks, including AMC, AIME, and OlympiadBench, show that our method improves reasoning accuracy over prior offline distillation algorithms and yields more stable reasoning traces while preserving instruction-following capabilities. Our work shows that lightweight, distribution-correction-aware training can substantially strengthen offline reasoning distillation without online rollouts.

cs.CL↗