arXiv subjects

Xiaoming Chen

Publications and source records attributed to Xiaoming Chen.

At least 37 records · Page 2Linked to original sources

arXiv2026

A First Exploration of Neuromorphic OT-CFM for Multi-Speaker VSR

Visual Speech Recognition (VSR) tasks in complex multi-speaker scenarios are severely hindered by rapid head motions, occlusions, and subtle lip articulations. Traditional RGB-based methods struggle here due to low rates and motion blur of frames. To overcome these, we propose LipsFlow, a neuromorphic-inspired VSR framework that converts RGB videos into high-temporal-resolution event streams. For multi-speaker, we employ ByteTrack tracking and TalkNet active speaker detection to temporally segment scenes into single-speaker clips, enabling focused per-speaker analysis. By explicitly capturing microsecond-level articulatory dynamics via learnable event-based representations, LipsFlow achieves inherent robustness against visual degradation. To efficiently model these dense event-based features and adapt to speaker-specific articulatory patterns, we introduce Optimal Transport Conditional Flow Matching (OT-CFM). It enforces deterministic, straight-line trajectory generation in a semantic latent space, slashing inference latency to just two Ordinary Differential Equation (ODE) steps. Furthermore, we design a Dual-Level Semantic Supervision mechanism combining token-level BERT weight tying and sentence-level priors to resolve homophene ambiguities. Validated on competitive benchmarks, LipsFlow achieves a state-of-the-art WER of 22.3\% at 240 ms latency, establishing a highly robust and efficient paradigm for event-based VSR.

cs.MM↗

arXiv2026

Can We Hear from Events? Generating Speech from Event Camera

Traditional RGB-based speech generation faces Temporal Granularity Mismatch since fixed camera exposure times inevitably blur the high-frequency articulatory transients essential for rendering emotional speech. To break this ceiling, we propose EventSpeech as a novel text-conditioned framework pioneering the use of neuromorphic events for expressive speech generation, since these microsecond-precise events naturally align with acoustic waveform dynamics. Our architecture integrates a dedicated Event Encoder to model sparse neuromorphic events alongside a multi-scale Audio Encoder featuring a Hierarchical Wavelet Contextualizer (HWC). A bidirectional alignment mechanism seamlessly synchronizes linguistic content and visual dynamics with dense acoustic features. Furthermore, we construct EVT-SPK as the first benchmark comprising large-scale synthetic data and real-world recordings from specialized neuromorphic hardware. Extensive evaluations demonstrate that EventSpeech significantly outperforms current baselines by preserving fine-grained emotions and resisting motion blur to establish a new paradigm for multimodal speech generation. Code and demo are available at https://xrfang-0102.github.io/EventSpeechWeb/.

cs.MM↗

arXiv2026

DWDP: Distributed Weight Data Parallelism for High-Performance LLM Inference on NVL72

Large language model (LLM) inference increasingly depends on multi-GPU execution, yet existing inference parallelization strategies require layer-wise inter-rank synchronization, making end-to-end performance sensitive to workload imbalance. We present DWDP (Distributed Weight Data Parallelism), an inference parallelization strategy that preserves data-parallel execution while offloading MoE weights across peer GPUs and fetching missing experts on demand. By removing collective inter-rank synchronization, DWDP allows each GPU to progress independently. We further address the practical overheads of this design with two optimizations for split-weight management and asynchronous remote-weight prefetch. Implemented in TensorRT-LLM and evaluated with DeepSeek-R1 on GB200 NVL72, DWDP improves end-to-end output TPS/GPU by 8.8% at comparable TPS/user in the 20-100 TPS/user serving range under 8K input sequence length and 1K output sequence length.

cs.DC↗

arXiv2026

Wireless large AI model: shaping the AI-empowered future of 6G and beyond

The emergence of sixth-generation and beyond communication systems is expected to fundamentally transform digital experiences through introducing unparalleled levels of intelligence, efficiency, and connectivity. A promising technology poised to enable this revolutionary vision is a wireless large AI model (WLAM), characterized by its exceptional capabilities in data processing, inference, and decision-making. In light of these remarkable capabilities, this paper provides a comprehensive survey of WLAM, explaining its fundamental principles, diverse applications, critical challenges, and future research opportunities. We begin by introducing the background of WLAM and analyzing the key synergies with wireless networks, emphasizing the mutual benefits. Subsequently, we explore the foundational characteristics of WLAM, delving into their unique relevance in wireless environments. Then, the role of WLAM in optimizing wireless communication systems across various use cases and the reciprocal benefits are systematically investigated. Furthermore, we discuss the integration of WLAM with emerging technologies, highlighting their potential to enable transformative capabilities and breakthroughs in wireless communication. Finally, we thoroughly examine the high-level challenges and discuss pivotal future research directions.

cs.IT↗

arXiv2026

On Systematic Performance of 3-D Holographic MIMO: Clarke, Kronecker, and 3GPP Models

Holographic multiple-input multiple-output (MIMO) has emerged as a key enabler for 6G networks, yet conventional planar implementations suffer from spatial correlation and mutual coupling at sub-wavelength spacing, which fundamentally limit the effective degrees of freedom (EDOF) and channel capacity. Three-dimensional (3-D) holographic MIMO offers a pathway to overcome these constraints by exploiting volumetric array configurations that enlarge the effective aperture and unlock additional spatial modes. This work presents the first systematic evaluation that jointly incorporates electromagnetic (EM) characteristics, such as mutual coupling and radiation efficiency, into the analysis of 3-D arrays under Clarke, Kronecker, and standardized 3rd Generation Partnership Project (3GPP) channel models. Analytical derivations and full-wave simulations demonstrate that 3-D architectures achieve higher EDOF, narrower beamwidths, and notable capacity improvements compared with planar baselines. In 3GPP urban macro channels with horizontal element spacing of 0.3 lambda, 3-D configurations yield approximately 20% capacity improvement over conventional 2-D arrays, confirming the robustness and scalability of volumetric designs under realistic conditions. These findings bridge the gap between theoretical feasibility and practical deployment, offering design guidance for next-generation 6G base station arrays.

eess.SP↗

arXiv2026

Guess-Verify-Refine: Data-Aware Top-K for Sparse-Attention Decoding on Blackwell via Temporal Correlation

Sparse-attention decoders rely on exact Top-K selection to choose the most important key-value entries for each query token. In long-context LLM serving, this Top-K stage runs once per decode query and becomes a meaningful latency bottleneck even when the indexer and attention kernels are already highly optimized. We present \textbf{Guess-Verify-Refine (GVR)}, a data-aware exact Top-K algorithm for sparse-attention decoding on NVIDIA Blackwell. GVR exploits temporal correlation across consecutive decode steps: it uses the previous step's Top-K as a prediction signal, computes pre-indexed statistics, narrows to a valid threshold by secant-style counting in 1-2 global passes, verifies candidates with a ballot-free collector, and finishes exact selection in shared memory. We connect this behavior to the Toeplitz / RoPE structure of DeepSeek Sparse Attention (DSA) indexer scores and validate the design on real DeepSeek-V3.2 workloads integrated into TensorRT-LLM. GVR achieves an average \textbf{1.88x} single-operator speedup over the production radix-select kernel, with up to \textbf{2.42x} per layer per step, while preserving bit-exact Top-K outputs. In controlled TEP8 min-latency deployment, it improves end-to-end TPOT by up to \textbf{7.52%} at 100K context, with larger gains at longer contexts and smaller but still positive gains under speculative decoding. While implemented and validated in the current TensorRT-LLM DSA stack on Blackwell, the same principle may extend to sparse-attention decoders whose decode-phase Top-K exhibits temporal stability.

cs.DC↗

arXiv2026

DynFOA: Generating First-Order Ambisonics with Conditional Diffusion for Dynamic and Acoustically Complex 360-Degree Videos

Spatial audio is crucial for immersive 360-degree video experiences, yet most 360-degree videos lack it due to the difficulty of capturing spatial audio during recording. Automatically generating spatial audio such as first-order ambisonics (FOA) from video therefore remains an important but challenging problem. In complex scenes, sound perception depends not only on sound source locations but also on scene geometry, materials, and dynamic interactions with the environment. However, existing approaches only rely on visual cues and fail to model dynamic sources and acoustic effects such as occlusion, reflections, and reverberation. To address these challenges, we propose DynFOA, a generative framework that synthesizes FOA from 360-degree videos by integrating dynamic scene reconstruction with conditional diffusion modeling. DynFOA analyzes the input video to detect and localize dynamic sound sources, estimate depth and semantics, and reconstruct scene geometry and materials using 3D Gaussian Splatting (3DGS). The reconstructed scene representation provides physically grounded features that capture acoustic interactions between sources, environment, and listener viewpoint. Conditioned on these features, a diffusion model generates spatial audio consistent with the scene dynamics and acoustic context. We introduce M2G-360, a dataset of 600 real-world clips divided into MoveSources, Multi-Source, and Geometry subsets for evaluating robustness under diverse conditions. Experiments show that DynFOA consistently outperforms existing methods in spatial accuracy, acoustic fidelity, distribution matching, and perceived immersive experience.

cs.SD↗

arXiv2026

HYLU: Hybrid Parallel Sparse LU Factorization

This article introduces HYLU, a hybrid parallel LU factorization-based general-purpose solver designed for efficiently solving sparse linear systems (Ax=b) on multi-core shared-memory architectures. The key technical feature of HYLU is the integration of hybrid numerical kernels so that it can adapt to various sparsity patterns of coefficient matrices. Tests on 37 sparse matrices from SuiteSparse Matrix Collection reveal that HYLU outperforms Intel MKL PARDISO in the numerical factorization phase by geometric means of 2.36X (for one-time solving) and 2.90X (for repeated solving). HYLU can be downloaded from https://github.com/chenxm1986/hylu.

cs.AR↗

arXiv2026

Less is More: Skim Transformer for Light Field Image Super-resolution

A light field image captures scenes through its micro-lens array, providing a rich representation that encompasses spatial and angular information. While this richness comes at significant data redundancy, most existing methods tend to indiscriminately utilize all the information from sub-aperture images (SAIs) in an attempt to harness every visual cue regardless of their disparity significance. However, this paradigm inevitably leads to disparity entanglement, a fundamental cause of inefficiency in light field image processing. To address this limitation, we introduce the Skim Transformer, a novel architecture inspired by the "less is more" philosophy. It features a multi-branch structure where each branch is dedicated to a specific disparity range by constructing its attention score matrix over a skimmed subset of SAIs, rather than all of them. Building upon it, we present SkimLFSR, an efficient yet powerful network for light field image super-resolution. Requiring only 67% of the prior leading method's parameters}, SkimLFSR achieves state-of-the-art results surpassing the best existing method by 0.63 dB and 0.35 dB PSNR at the 2x and 4x tasks, respectively. Through in-depth analyses, we reveal that SkimLFSR, guided by the predefined skimmed SAI sets as prior knowledge, demonstrates distinct disparity-aware behaviors in attending to visual cues. Last but not least, we conduct an experiment to validate SkimLFSR's generalizability across different angular resolutions, where it achieves competitive performance on a larger angular resolution without any retraining or major network modifications. These findings highlight its effectiveness and adaptability as a promising paradigm for light field image processing.

eess.IV↗

arXiv2026

A Quantitative Method for Evaluating Security Boundaries in Quantum Key Distribution Combined with Block Ciphers

With the rapid development of quantum computing, classical cryptography systems are increasingly vulnerable to security threats, thereby highlighting the urgency of constructing architectures that are resilient to quantum computing attacks. While Quantum Key Distribution (QKD) offers security with information-theoretic guarantees, its relatively low key generation rate necessitates integration with classical cryptographic techniques, particularly block ciphers such as AES and SM4, to facilitate practical applications. However, when a single QKD-key is employed to encrypt multiple data blocks, the reduction in cryptographic security strength has not yet been quantitatively analyzed. In this work, we focus on the security strength in the application scenario where QKD is combined with block ciphers. We propose a quantitative evaluation method for the security benefits of the QKD-key renewal period, aiming to provide a precise measure of the cryptographic security strength in such hybrid systems. Our method is based on concrete security paradigm of block cipher modes of operation. We demonstrate that under practical security level requirements, for files consisting of specific blocks, rekeying k times can provide an additional log2(k) to 2log2(k) bits of security. Our research offers a novel perspective on balancing the security and efficiency of QKD-based encryption.

cs.CR↗

arXiv2026

ICWLM: A Multi-Task Wireless Large Model via In-Context Learning

The rapid evolution of wireless communication technologies, particularly massive multiple-input multiple-output (mMIMO) and millimeter-wave (mmWave), introduces significant network complexity and computational demands. Significant research efforts have been made to improve physical layer performance by resorting to deep learning (DL) methods, which, however, are usually task-specific and struggle with data scarcity and generalization. To address these challenges, we propose a novel In-Context Wireless Large Model (ICWLM), a wireless-native foundation model designed for simultaneous multi-task learning at the physical layer. Unlike conventional methods that adapt wireless data to pre-trained large language models (LLMs), ICWLM is trained directly on large-scale, mixed wireless datasets from scratch. It jointly solves multiple classical physical layer problems, including multi-user precoding (sum-rate maximization and max-min SINR) and channel prediction. A key innovation of ICWLM is its utilization of in-context learning (ICL), enabling the model to adapt to varying system configurations and channel conditions with minimal demonstration pairs, eliminating the need for extensive retraining. Extensive simulation results demonstrate that ICWLM achieves competitive performance compared to task-specific methods while exhibiting remarkable generalization capabilities to unseen system configurations. This work offers a promising paradigm for developing unified and adaptive AI models for future wireless networks, potentially reducing deployment complexity and enhancing intelligent resource management.

eess.SP↗

arXiv2025

SketchPlay: Intuitive Creation of Physically Realistic VR Content with Gesture-Driven Sketching

Creating physically realistic content in VR often requires complex modeling tools or predefined 3D models, textures, and animations, which present significant barriers for non-expert users. In this paper, we propose SketchPlay, a novel VR interaction framework that transforms humans' air-drawn sketches and gestures into dynamic, physically realistic scenes, making content creation intuitive and playful like drawing. Specifically, sketches capture the structure and spatial arrangement of objects and scenes, while gestures convey physical cues such as velocity, direction, and force that define movement and behavior. By combining these complementary forms of input, SketchPlay captures both the structure and dynamics of user-created content, enabling the generation of a wide range of complex physical phenomena, such as rigid body motion, elastic deformation, and cloth dynamics. Experimental results demonstrate that, compared to traditional text-driven methods, SketchPlay offers significant advantages in expressiveness, and user experience. By providing an intuitive and engaging creation process, SketchPlay lowers the entry barrier for non-expert users and shows strong potential for applications in education, art, and immersive storytelling.

cs.HC↗

arXiv2025

Beyond Subspace Isolation: Many-to-Many Transformer for Light Field Image Super-resolution

The effective extraction of spatial-angular features plays a crucial role in light field image super-resolution (LFSR) tasks, and the introduction of convolution and Transformers leads to significant improvement in this area. Nevertheless, due to the large 4D data volume of light field images, many existing methods opted to decompose the data into a number of lower-dimensional subspaces and perform Transformers in each sub-space individually. As a side effect, these methods inadvertently restrict the self-attention mechanisms to a One-to-One scheme accessing only a limited subset of LF data, explicitly preventing comprehensive optimization on all spatial and angular cues. In this paper, we identify this limitation as subspace isolation and introduce a novel Many-to-Many Transformer (M2MT) to address it. M2MT aggregates angular information in the spatial subspace before performing the self-attention mechanism. It enables complete access to all information across all sub-aperture images (SAIs) in a light field image. Consequently, M2MT is enabled to comprehensively capture long-range correlation dependencies. With M2MT as the foundational component, we develop a simple yet effective M2MT network for LFSR. Our experimental results demonstrate that M2MT achieves state-of-the-art performance across various public datasets, and it offers a favorable balance between model performance and efficiency, yielding higher-quality LFSR results with substantially lower demand for memory and computation. We further conduct in-depth analysis using local attribution maps (LAM) to obtain visual interpretability, and the results validate that M2MT is empowered with a truly non-local context in both spatial and angular subspaces to mitigate subspace isolation and acquire effective spatial-angular representation.

eess.IV↗

arXiv2025

Integration of Navigation and Remote Sensing in LEO Satellite Constellations

Low earth orbit (LEO) satellite constellations are becoming a cornerstone of next-generation satellite networks, enabling worldwide high-precision navigation and high-quality remote sensing. This paper proposes a novel dual-function LEO satellite constellation frame structure that effectively integrating navigation and remote sensing. Then, the Cramer-Rao bound (CRB)-based positioning, velocity measurement, and timing (PVT) error and the signal-to-ambiguity-interference-noise ratio (SAINR) are derived as performance metrics for navigation and remote sensing, respectively. Based on it, a joint beamforming design is proposed by minimizing the average weighted PVT error for navigation user equipments (UEs) while ensuring SAINR requirement for remote sensing. Simulation results validate the proposed multi-satellite cooperative beamforming design, demonstrating its effectiveness as an integrated solution for next-generation multi-function LEO satellite constellations.

cs.IT↗

arXiv2025

NVS-SQA: Exploring Self-Supervised Quality Representation Learning for Neurally Synthesized Scenes without References

Neural View Synthesis (NVS), such as NeRF and 3D Gaussian Splatting, effectively creates photorealistic scenes from sparse viewpoints, typically evaluated by quality assessment methods like PSNR, SSIM, and LPIPS. However, these full-reference methods, which compare synthesized views to reference views, may not fully capture the perceptual quality of neurally synthesized scenes (NSS), particularly due to the limited availability of dense reference views. Furthermore, the challenges in acquiring human perceptual labels hinder the creation of extensive labeled datasets, risking model overfitting and reduced generalizability. To address these issues, we propose NVS-SQA, a NSS quality assessment method to learn no-reference quality representations through self-supervision without reliance on human labels. Traditional self-supervised learning predominantly relies on the "same instance, similar representation" assumption and extensive datasets. However, given that these conditions do not apply in NSS quality assessment, we employ heuristic cues and quality scores as learning objectives, along with a specialized contrastive pair preparation process to improve the effectiveness and efficiency of learning. The results show that NVS-SQA outperforms 17 no-reference methods by a large margin (i.e., on average 109.5% in SRCC, 98.6% in PLCC, and 91.5% in KRCC over the second best) and even exceeds 16 full-reference methods across all evaluation metrics (i.e., 22.9% in SRCC, 19.1% in PLCC, and 18.6% in KRCC over the second best).

cs.CV↗

arXiv2025

V-GameGym: Visual Game Generation for Code Large Language Models

Code large language models have demonstrated remarkable capabilities in programming tasks, yet current benchmarks primarily focus on single modality rather than visual game development. Most existing code-related benchmarks evaluate syntax correctness and execution accuracy, overlooking critical game-specific metrics such as playability, visual aesthetics, and user engagement that are essential for real-world deployment. To address the gap between current LLM capabilities in algorithmic problem-solving and competitive programming versus the comprehensive requirements of practical game development, we present V-GameGym, a comprehensive benchmark comprising 2,219 high-quality samples across 100 thematic clusters derived from real-world repositories, adopting a novel clustering-based curation methodology to ensure both diversity and structural completeness. Further, we introduce a multimodal evaluation framework with an automated LLM-driven pipeline for visual code synthesis using complete UI sandbox environments. Our extensive analysis reveals that V-GameGym effectively bridges the gap between code generation accuracy and practical game development workflows, providing quantifiable quality metrics for visual programming and interactive element generation.

cs.SE↗

arXiv2025

Design of RIS-UAV-Assisted LEO Satellite Constellation Communication

Low Earth orbit (LEO) satellite constellations play a pivotal role in sixth-generation (6G) wireless networks by providing global coverage, massive connections, and huge capacity. In this paper, we present a novel LEO satellite constellation communication framework, where a reconfigurable intelligent surface-mounted unmanned aerial vehicle (RIS-UAV) is deployed to improve the communication quality of multiple terrestrial user equipments (UEs) under the condition of long distance between satellite and ground. To reduce the overhead for channel state information (CSI) acquisition with multiple-satellite collaboration, statistical CSI (sCSI) is utilized in the system. In such a situation, we first derive an approximated but exact expression for ergodic rate of each UE. Then, we aim to maximize the minimum approximated UE ergodic rate by the proposed alternating optimization (AO)-based algorithm that jointly optimizes LEO satellite beamforming, RIS phase shift, and UAV trajectory. Finally, extensive simulations are conducted to demonstrate the superiority of the proposed algorithm in terms of spectrum efficiency over baseline algorithms.

cs.IT↗

arXiv2025

QoS-Driven Satellite Constellation Design for LEO Satellite Internet of Things

Low Earth orbit (LEO) satellite Internet of Things (IoT) has been identified as one of the important components of the sixth-generation (6G) non-terrestrial networks (NTN) to provide ubiquitous connectivity. Due to the low orbit altitude and high mobility, a massive number of satellites are required to form a global continuous coverage constellation, leading to a high construction cost. To this end, this paper proposes a LEO satellite IoT constellation design algorithm with the goal of minimizing the total cost while satisfying quality of service (QoS) requirements in terms of coverage ratio and communication quality. Specifically, with a novel fitness function and efficient algorithm's operators, the proposed algorithm converges more quickly and achieves lower constellation construction cost compared to baseline algorithms under the same QoS requirements. Theoretical analysis proves the global and fast convergence of the proposed algorithm due to a novel fitness function. Finally, extensive simulation results confirm the effectiveness of the proposed algorithm in LEO satellite IoT constellation design.

cs.IT↗