arXiv subjects

Wei Liang

Publications and source records attributed to Wei Liang.

At least 19 recordsLinked to original sources

arXiv2026

Scaling phoneme-based TTS augmentation for ASR: A unified pipeline and controlled study

Synthetic speech offers scalable supervision for automatic speech recognition (ASR), but its benefit depends on text selection, reference speech, and augmentation scale. We present a phoneme-based TTS-to-ASR pipeline using a single TTS model jointly trained from scratch on Arabic, French, Italian, and Portuguese with the F5-TTS architecture and language-monolingual ASR systems cover 13 test sets. Across the synthesis-scale sweep, random augmentation improves over matched real-only continuation on 11 sets. In the selection comparison, PFGS improves over real-only training on 12 sets and over random selection on nine, with a maximum relative WER reduction of 19.3% against random selection. With target texts and synthesis counts fixed, reference-speech filtering reduces absolute WER by 0.29 and 0.59 points on Italian and French Common Voice, respectively. These findings support treating TTS augmentation as a synthetic-corpus construction problem, rather than merely a question of generation scale.

cs.CL↗

arXiv2026

Embodied Scene Rearrangement Planning

This paper introduces Embodied Scene Rearrangement Planning (ESRP), a novel task requiring embodied agents to rearrange furniture in 3D scenes to match a target configuration using only egocentric observations and a top-down target layout. Unlike prior rearrangement tasks, ESRP precludes global state access and introduces mutual object occlusions, reflecting the practical constraints of real-world robotic deployment. These factors make aligning partial egocentric observations with the global target layout particularly challenging for long-horizon planning. To facilitate research, we present ESRP-Bench, a comprehensive benchmark built on OmniGibson featuring over 5,400 scene pairs and 8,200 objects. We define three multi-level metrics to evaluate rearrangement quality and provide four baselines: a hierarchical task-and-motion planning method, a vision-language-model-based method, and two learning-based approaches (IL and RL). Experimental results demonstrate that current methods struggle to complete the task efficiently, highlighting ESRP as a challenging frontier for embodied agents in scene understanding and long-horizon task planning. This work serves as a stepping stone toward deploying intelligent agents in real-world scenarios. Project page: https://pie-lab.cn/ESRP/.

cs.RO↗

arXiv2026

SONG: A Photorealistic 3D Gaussian Simulation Platform for Benchmarking Social Navigation

Social navigation has progressed from simplified 2D environments toward a more general vision-based setting, in which a robot needs to achieve socially compliant behavior purely from onboard visual observations. Yet supporting simulation platforms have not kept pace: existing options either lack visual observations, lack moving human avatars, or fall short of real-world fidelity in appearance and pedestrian behavior, offering limited support for advancing vision-based social navigation. We introduce SONG, a SOcial Navigation platform powered by 3D Gaussian splatting (3DGS). It leverages 3DGS for both scene and avatar representations, drives pedestrians using semantically grounded trajectories generated by a large language model, and synthesizes their full-body motion with a trajectory-conditioned generator to produce continuous, natural movement. On top of the platform, we curate SONG-Bench, a set of evaluation episodes stratified by difficulty, and propose a multi-dimensional metric suite covering effectiveness, safety, and social compliance. A systematic evaluation of representative navigation baselines reveals three findings: (a) vision-based social navigation is far from solved; (b) a critical safety deficit precedes social etiquette; (c) real-world data matters more than model scale. Crucially, we demonstrate that fine-tuning on our curated data effectively improves the success rate in real-world environments. We hope our platform provides a faithful and rigorous testbed for the next generation of vision-based social navigation research.

cs.RO↗

arXiv2026

Diarization-Guided Qwen-ASR Adaptation for Multilingual Two-Speaker Conversational Speech

This paper describes our self-designed system for Task 1 of the MLC-SLM 2026 Challenge for multilingual two-speaker conversational speech. The system combines a modular speaker diarization front end with a challenge-adapted Qwen3-ASR-1.7B recognizer. The diarization front end performs voice activity detection, subsegment generation, CAMPPlus speaker embedding extraction, two-speaker spectral clustering, and RTTM-based audio segmentation. The resulting speaker-attributed segments are grouped by language or region and decoded by the adapted ASR model. For ASR adaptation, we first perform supervised full fine-tuning on the official training data, then apply LoRA fine-tuning with synthetic speech generated by a three-pipeline TTS-based synthetic speech augmentation framework, and finally refine the model using GRPO reinforcement learning with rewards based on WER/CER and penalties for hallucination, repetition, and length deviation. On the official development set, the full system achieves an average tcpMER of 23.70, reducing the error rate by 6.83 absolute points relative to the released Qwen-ASR-1.7B performance. On the final evaluation set, the system achieves an average tcpMER of 17.97. Ablation results show that supervised fine-tuning provides the largest gain, while synthetic-speech LoRA adaptation and reinforcement learning further improve robustness.

cs.CL↗

arXiv2026

FloVerse: Floor Plan-Guided Multi-Modal Navigation

Floor plans encapsulate compact spatial priors, enabling agents to navigate unseen scenes more efficiently. While prior work has explored floor plan-guided navigation, it has focused mainly on PointNav and a limited set of environments. To bridge this gap, we introduce FloVerse, a new task for floor plan-guided embodied navigation that unifies PointNav, ObjectNav, and ImageNav. To support FloVerse, we assemble FloVerse-1.6K, a large-scale dataset of 1.6K scenes from HM3D and Gibson 4+, paired with corresponding floor plans, comprising 240K expert trajectories and 12M RGBD frames. We further propose ThreeDiff, a two-stage imitation learning policy comprising a planner, a diffusion-based multimodal goal-reasoning module trained via masked-modality modeling, and a refiner, a depth-based trajectory-refinement module for safe execution. Extensive experiments demonstrate that (1) floor-plan priors improve navigation performance across all goal modalities, and (2) ThreeDiff implicitly captures spatial information from floor plans. These results underscore the effectiveness of spatial priors and validate our proposed unified approach for floor plan-guided embodied navigation.

cs.RO↗

arXiv2026

OpenSkill: Open-World Self-Evolution for LLM Agents

Self-evolving agents requires adaptation after deployment, but existing approaches assume a usable learning loop, such as curated skills, successful trajectories, or verifier signals. Real open-world deployments may provide none of these, offering only a task prompt. In this work, we study open-world self-evolution, where an agent must build both its skills and its own verification signals from scratch, using open-world resources but no target-task supervision. We propose OpenSkill, a framework that bootstraps this loop: it acquires grounded knowledge and verification anchors from documentation, repositories, and the web, synthesizes them into transferable skills, and refines those skills against self-built virtual tasks grounded in the anchors rather than in target answers. The open world thus supplies both the knowledge to be learned and a supervision-independent practice environment, with target-task supervision reserved for final evaluation. Across three benchmarks and two target agents, OpenSkill attains the best automated pass rate while satisfying the no-supervision constraint. Analysis shows its skills transfer across models without model-specific adaptation, and its self-built verifier aligns with ground-truth outcomes despite never accessing them.

cs.AI↗

arXiv2026

SDE-Driven Spatio-Temporal Hypergraph Neural Networks for Irregular Longitudinal fMRI Connectome Modeling in Alzheimer's Disease

Longitudinal neuroimaging is essential for modeling disease progression in Alzheimer's disease (AD), yet irregular sampling and missing visits pose substantial challenges for learning reliable temporal representations. To address this challenge, we propose SDE-HGNN, a stochastic differential equation (SDE)-driven spatio-temporal hypergraph neural network for irregular longitudinal fMRI connectome modeling. The framework first employs an SDE-based reconstruction module to recover continuous latent trajectories from irregular observations. Based on these reconstructed representations, dynamic hypergraphs are constructed to capture higher-order interactions among brain regions over time. To further model temporal evolution, hypergraph convolution parameters evolve through SDE-controlled recurrent dynamics conditioned on inter-visit intervals, enabling disease-stage-adaptive connectivity modeling. We also incorporate a sparsity-based importance learning mechanism to identify salient brain regions and discriminative connectivity patterns. Extensive experiments on the OASIS-3 and ADNI cohorts demonstrate consistent improvements over state-of-the-art graph and hypergraph baselines in AD progression prediction. The source code is available at https://anonymous.4open.science/r/SDE-HGNN-017F.

cs.LG↗

arXiv2026

Investigation of the $D^{0} \rightarrow K_S^{0} \pi^{0} \eta,\ K_S^{0} \pi^{0} \pi^0$ decays

Inspired by the invariant mass distributions for the decays $D^{0} \rightarrow K_S^{0} \pi^{0} \eta$ and $D^{0} \rightarrow K_S^{0} \pi^{0} \pi^{0}$ reported by the BESIII Collaboration, we investigate these processes with an unified final state interaction formalism by incorporating both the $S$-wave pseudoscalar meson-pseudoscalar meson interactions within a chiral unitary approach and the $P$-wave contribution from the intermediate resonance $\bar{K}^*(892)$. By performing a combined fit to the invariant mass spectra and taking into account the coherence between the $S$- and $P$-waves, our results are in agreement with the experimental data. For the decay $D^{0} \rightarrow K_S^{0} \pi^{0} \eta$, the structure near 1.0 GeV in the $\pi^0 \eta$ invariant mass distribution corresponds to the signal of the $a_{0}(980)$, which is dynamically generated from the $S$-wave interactions. In the case of $D^{0} \rightarrow K_S^{0} \pi^{0} \pi^{0}$, the near-threshold enhancement in the $\pi^0\pi^0$ mass distribution arises from the combined contributions of the $f_0(500)$ and the intermediate $\bar{K}^*(892)$, while the cusp-like structure around 1 GeV$^2$ is associated with the $f_0(980)$.

hep-ph↗

arXiv2026

OmniClone: Engineering a Robust, All-Rounder Whole-Body Humanoid Teleoperation System

Whole-body humanoid teleoperation enables humans to remotely control humanoid robots, serving as both a real-time operational tool and a scalable engine for collecting demonstrations for autonomous learning. Despite recent advances, existing systems are validated using aggregate metrics that conflate distinct motion regimes, masking critical failure modes. This lack of diagnostic granularity, compounded by tightly coupled and labor-intensive system configurations, hinders robust real-world deployment. A key open challenge is building a teleoperation system that is simultaneously robust, versatile, and affordable for practical use. Here we present OmniClone, a whole-body humanoid teleoperation system that achieves high-fidelity, multi-skill control on a single consumer GPU with modest data requirements. Central to our approach is OmniBench, a diagnostic benchmark that evaluates policies across stratified motion categories and difficulty levels on unseen motions, exposing the narrow specialization of prior systems. Guided by these diagnostics, we identify an optimized training data recipe and integrate system-level improvements: subject-agnostic retargeting and robust communication, that collectively reduce Mean Per-Joint Position Error (MPJPE) by over 66% while requiring orders-of-magnitude fewer computational resources than comparable methods. Crucially, OmniClone is control-source-agnostic: a single unified policy supports real-time teleoperation, generated motion playback, and Vision-Language-Action (VLA) models, while generalizing across operators of vastly different body proportions. By uniting diagnostic evaluation with practical engineering, OmniClone provides an accessible foundation for scalable humanoid teleoperation and autonomous learning.

cs.RO↗

arXiv2026

Joint Imaging-ROI Representation Learning via Cross-View Contrastive Alignment for Brain Disorder Classification

Brain imaging classification is commonly approached from two perspectives: modeling the full image volume to capture global anatomical context, or constructing ROI-based graphs to encode localized and topological interactions. Although both representations have demonstrated independent efficacy, their relative contributions and potential complementarity remain insufficiently understood. Existing fusion approaches are typically task-specific and do not enable controlled evaluation of each representation under consistent training settings. To address this gap, we propose a unified cross-view contrastive framework for joint imaging-ROI representation learning. Our method learns subject-level global (imaging) and local (ROI-graph) embeddings and aligns them in a shared latent space using a bidirectional contrastive objective, encouraging representations from the same subject to converge while separating those from different subjects. This alignment produces comparable embeddings suitable for downstream fusion and enables systematic evaluation of imaging-only, ROI-only, and joint configurations within a unified training protocol. Extensive experiments on the ADHD-200 and ABIDE datasets demonstrate that joint learning consistently improves classification performance over either branch alone across multiple backbone choices. Moreover, interpretability analyses reveal that imaging-based and ROI-based branches emphasize distinct yet complementary discriminative patterns, explaining the observed performance gains. These findings provide principled evidence that explicitly integrating global volumetric and ROI-level representations is a promising direction for neuroimaging-based brain disorder classification. The source code is available at https://anonymous.4open.science/r/imaging-roi-contrastive-152C/.

cs.CV↗

arXiv2026

Abstraction Generation for Generalized Planning with Pretrained Large Language Models

Qualitative Numerical Planning (QNP) serves as an important abstraction model for generalized planning (GP), which aims to compute general plans that solve multiple instances at once. Recent works show that large language models (LLMs) can function as generalized planners. This work investigates whether LLMs can serve as QNP abstraction generators for GP problems and how to fix abstractions via automated debugging. We propose a prompt protocol: input a GP domain and training tasks to LLMs, prompting them to generate abstract features and further abstract the initial state, action set, and goal into QNP problems. An automated debugging method is designed to detect abstraction errors, guiding LLMs to fix abstractions. Experiments demonstrate that under properly guided by automated debugging, some LLMs can generate useful QNP abstractions.

cs.AI↗

arXiv2026

Optically locked low-noise photonic microwave oscillator

The next-generation sensing and communication applications rely on high-frequency microwave generation with low-noise. The microwave photonic technology is promising by the practical application is limited by its complex architecture so far. Here, we demonstrate an optically locked low-noise photonic microwave oscillator, so that all the optical components are packaged within a small module of 166 mL, and low noise microwave generation is achieved at 10.4 GHz with single-sideband phase noise of -54 dBc/Hz at 10 Hz, -141 dBc/Hz at 10 kHz, and -162 dBc/Hz at 10 MHz offset. Above performance arises from a dual-laser self-injection-locking scheme to a single Fabry-Perot cavity with high Q exceeding 10^8, with over 20 dB common-mode noise suppression. The low-noise nature of such reference is coherently transferred to the X-band through a high-performance TFLN electro-optic comb chip, thereby overcoming long-standing barriers in photonic microwave integration to enable truly field-deployable low-noise microwave generation.

physics.optics↗

arXiv2026

Audio-Driven Talking Face Generation with Blink Embedding and Hash Grid Landmarks Encoding

Dynamic Neural Radiance Fields (NeRF) have demonstrated considerable success in generating high-fidelity 3D models of talking portraits. Despite significant advancements in the rendering speed and generation quality, challenges persist in accurately and efficiently capturing mouth movements in talking portraits. To tackle this challenge, we propose an automatic method based on blink embedding and hash grid landmarks encoding in this study, which can substantially enhance the fidelity of talking faces. Specifically, we leverage facial features encoded as conditional features and integrate audio features as residual terms into our model through a Dynamic Landmark Transformer. Furthermore, we employ neural radiance fields to model the entire face, resulting in a lifelike face representation. Experimental evaluations have validated the superiority of our approach to existing methods.

cs.CV↗

arXiv2026

SelfieAvatar: Real-time Head Avatar reenactment from a Selfie Video

Head avatar reenactment focuses on creating animatable personal avatars from monocular videos, serving as a foundational element for applications like social signal understanding, gaming, human-machine interaction, and computer vision. Recent advances in 3D Morphable Model (3DMM)-based facial reconstruction methods have achieved remarkable high-fidelity face estimation. However, on the one hand, they struggle to capture the entire head, including non-facial regions and background details in real time, which is an essential aspect for producing realistic, high-fidelity head avatars. On the other hand, recent approaches leveraging generative adversarial networks (GANs) for head avatar generation from videos can achieve high-quality reenactments but encounter limitations in reproducing fine-grained head details, such as wrinkles and hair textures. In addition, existing methods generally rely on a large amount of training data, and rarely focus on using only a simple selfie video to achieve avatar reenactment. To address these challenges, this study introduces a method for detailed head avatar reenactment using a selfie video. The approach combines 3DMMs with a StyleGAN-based generator. A detailed reconstruction model is proposed, incorporating mixed loss functions for foreground reconstruction and avatar image generation during adversarial training to recover high-frequency details. Qualitative and quantitative evaluations on self-reenactment and cross-reenactment tasks demonstrate that the proposed method achieves superior head avatar reconstruction with rich and intricate textures compared to existing approaches.

cs.CV↗

arXiv2025

The Wiener Path Integral Interpretation of the 3:1 Combat Rule

The Wiener path integral framework is proposed to model military combat dynamics by incorporating the neglected stochastic effects to the Lanchester's square law. This framework is applied to evaluate the empirical 3:1 combat rule, which posits that an attacker requires a threefold force superiority to achieve victory. Specifically, the attacker's winning probability is computed utilizing a semi-analytical Rayleigh-Ritz method. Numerical results demonstrate that the validity of the rule critically depends on specific parameter regimes, primarily contingent upon the relative combat effectiveness ratio between the opposing forces and the tolerance for attrition. This work establishes a physics-informed theoretical bridge between statistical mechanics and military operations research for analyzing uncertain combat systems.

physics.soc-ph↗

arXiv2025

Human Motion Estimation with Everyday Wearables

While on-body device-based human motion estimation is crucial for applications such as XR interaction, existing methods often suffer from poor wearability, expensive hardware, and cumbersome calibration, which hinder their adoption in daily life. To address these challenges, we present EveryWear, a lightweight and practical human motion capture approach based entirely on everyday wearables: a smartphone, smartwatch, earbuds, and smart glasses equipped with one forward-facing and two downward-facing cameras, requiring no explicit calibration before use. We introduce Ego-Elec, a 9-hour real-world dataset covering 56 daily activities across 17 diverse indoor and outdoor environments, with ground-truth 3D annotations provided by the motion capture (MoCap), to facilitate robust research and benchmarking in this direction. Our approach employs a multimodal teacher-student framework that integrates visual cues from egocentric cameras with inertial signals from consumer devices. By training directly on real-world data rather than synthetic data, our model effectively eliminates the sim-to-real gap that constrains prior work. Experiments demonstrate that our method outperforms baseline models, validating its effectiveness for practical full-body motion estimation.

cs.CV↗

arXiv2025

The Least Action-Augmented Lanchester Model

The principle of least action, a fundamental principle in variational mechanics with broad applicability to classical physical systems, is employed to formulate a novel attrition model for combat dynamics. This formulation extends the Lanchester's square law through second-order temporal derivatives by requiring the resultant Euler-Lagrange equation to coincide with the classical Lanchester's equation. Initial conditions at a specified temporal point enable determination of subsequent system evolution through action minimization, while terminal boundary conditions permit backward reconstruction of combat trajectories. The model's validity is examined through historical analysis of WWII engagements: the Battle of Kursk and the Battle of Iwo Jima. Comparative studies with conventional Lanchester's square models demonstrate marked improvements in predictive accuracy regarding force strength progression, particularly in capturing non-linear attrition patterns characteristic of prolonged engagements.

physics.soc-ph↗

arXiv2025

TFCDiff: Robust ECG Denoising via Time-Frequency Complementary Diffusion

Ambulatory electrocardiogram (ECG) readings are prone to mixed noise from physical activities, including baseline wander (BW), muscle artifact (MA), and electrode motion artifact (EM). Developing a method to remove such complex noise and reconstruct high-fidelity signals is clinically valuable for diagnostic accuracy. However, denoising of multi-beat ECG segments remains understudied and poses technical challenges. To address this, we propose Time-Frequency Complementary Diffusion (TFCDiff), a novel approach that operates in the Discrete Cosine Transform (DCT) domain and uses the DCT coefficients of noisy signals as conditioning input. To refine waveform details, we incorporate Temporal Feature Enhancement Mechanism (TFEM) to reinforce temporal representations and preserve key physiological information. Comparative experiments on a synthesized dataset demonstrate that TFCDiff achieves state-of-the-art performance across five evaluation metrics. Furthermore, TFCDiff shows superior generalization on the unseen SimEMG Database, outperforming all benchmark models. Notably, TFCDiff processes raw 10-second sequences and maintains robustness under flexible random mixed noise (fRMN), enabling plug-and-play deployment in wearable ECG monitors for high-motion scenarios. Source code is available at https://github.com/Miroircivil/TFCDiff.

eess.SP↗