Searcharxiv⌕ Search

arXiv subjects

Lei Yang

Publications and source records attributed to Lei Yang.

At least 109 records · Page 6Linked to original sources
arXiv2026

EmbodMocap: In-the-Wild 4D Human-Scene Reconstruction for Embodied Agents

Human behaviors in the real world naturally encode rich, long-term contextual information that can be leveraged to train embodied agents for perception, understanding, and acting. However, existing capture systems typically rely on costly studio setups and wearable devices, limiting the large-scale collection of scene-conditioned human motion data in the wild. To address this, we propose EmbodMocap, a portable and affordable data collection pipeline using two moving iPhones. Our key idea is to jointly calibrate dual RGB-D sequences to reconstruct both humans and scenes within a unified metric world coordinate frame. The proposed method allows metric-scale and scene-consistent capture in everyday environments without static cameras or markers, bridging human motion and scene geometry seamlessly. Compared with optical capture ground truth, we demonstrate that the dual-view setting exhibits a remarkable ability to mitigate depth ambiguity, achieving superior alignment and reconstruction performance over single iphone or monocular models. Based on the collected data, we empower three embodied AI tasks: monocular human-scene-reconstruction, where we fine-tune on feedforward models that output metric-scale, world-space aligned humans and scenes; physics-based character animation, where we prove our data could be used to scale human-object interaction skills and scene-aware motion tracking; and robot motion control, where we train a humanoid robot via sim-to-real RL to replicate human motions depicted in videos. Experimental results validate the effectiveness of our pipeline and its contributions towards advancing embodied AI research.

cs.CV↗
arXiv2026

Calibrating optical galaxy cluster projection effects with sparse spectroscopic samples: A clustering redshift approach

Wide-field optical imaging surveys are efficient at identifying galaxy clusters, but optically identified clusters suffer from projection effects--physically unassociated galaxies along the line of sight can be misidentified as cluster members due to distance uncertainties. Previous studies have used spectroscopic follow-up observations of cluster members to quantify projection effects; however, such follow-up efforts cannot keep pace with the rapidly growing cluster samples. On the other hand, spectroscopic surveys designed for large-scale structure studies collect tens of millions of spectra but tend to have sparse spectra in cluster regions. To bridge this gap, we develop a clustering redshift approach that cross-correlates cluster members with sparse, non-cluster-targeted spectroscopic galaxy samples. We validate this approach using the Cardinal simulation, recovering the correct spectroscopic distribution and projection effect parameters of redMaPPer cluster members. Our approach is insensitive to the selection of the spectroscopic sample and paves the way for calibrating the upcoming LSST clusters using DESI and Roman spectroscopic samples.

astro-ph.CO↗
arXiv2026

Generalizable Detection of AI Generated Images with Large Models and Fuzzy Decision Tree

The malicious use and widespread dissemination of AI-generated images pose a serious threat to the authenticity of digital content. Existing detection methods exploit low-level artifacts left by common manipulation steps within the generation pipeline, but they often lack generalization due to model-specific overfitting. Recently, researchers have resorted to Multimodal Large Language Models (MLLMs) for AIGC detection, leveraging their high-level semantic reasoning and broad generalization capabilities. While promising, MLLMs lack the fine-grained perceptual sensitivity to subtle generation artifacts, making them inadequate as standalone detectors. To address this issue, we propose a novel AI-generated image detection framework that synergistically integrates lightweight artifact-aware detectors with MLLMs via a fuzzy decision tree. The decision tree treats the outputs of basic detectors as fuzzy membership values, enabling adaptive fusion of complementary cues from semantic and perceptual perspectives. Extensive experiments demonstrate that the proposed method achieves state-of-the-art accuracy and strong generalization across diverse generative models.

cs.CV↗
arXiv2026

The Quest for Generalizable Motion Generation: Data, Model, and Evaluation

Despite recent advances in 3D human motion generation (MoGen) on standard benchmarks, existing text-to-motion models still face a fundamental bottleneck in their generalization capability. In contrast, adjacent generative fields, most notably video generation (ViGen), have demonstrated remarkable generalization in modeling human behaviors, highlighting transferable insights that MoGen can leverage. Motivated by this observation, we present a comprehensive framework that systematically transfers knowledge from ViGen to MoGen across three key pillars: data, modeling, and evaluation. First, we introduce ViMoGen-228K, a large-scale dataset comprising 228,000 high-quality motion samples that integrates high-fidelity optical MoCap data with semantically annotated motions from web videos and synthesized samples generated by state-of-the-art ViGen models. The dataset includes both text-motion pairs and text-video-motion triplets, substantially expanding semantic diversity. Second, we propose ViMoGen, a flow-matching-based diffusion transformer that unifies priors from MoCap data and ViGen models through gated multimodal conditioning. To enhance efficiency, we further develop ViMoGen-light, a distilled variant that eliminates video generation dependencies while preserving strong generalization. Finally, we present MBench, a hierarchical benchmark designed for fine-grained evaluation across motion quality, prompt fidelity, and generalization ability. Extensive experiments show that our framework significantly outperforms existing approaches in both automatic and human evaluations. The code, data, and benchmark will be made publicly available. Homepage: https://motrixlab.github.io/2026_iclr_vimogen.

cs.CV↗
arXiv2026

Convergence Analysis of a Relative-type Inexact Preconditioned Proximal ALM for Convex Nonlinear Programming

This article investigates the convergence properties of a relative-type inexact preconditioned proximal augmented Lagrangian method (rip$^2$ALM) for convex nonlinear programming, a fundamental class of optimization problems with broad applications in science and engineering. Inexact proximal augmented Lagrangian methods have proven to be highly effective for solving such problems, owing to their attractive theoretical properties and strong practical performance. However, the convergence behavior of the relative-type inexact preconditioned variant remains insufficiently understood. This work aims to reduce this gap by rigorously establishing the global convergence of the sequence generated by rip$^2$ALM and proving its asymptotic (super)linear convergence rate under standard assumptions. In addition, we derive the global ergodic convergence rate with respect to both the primal feasibility violation and the primal objective residual, thereby offering a more comprehensive understanding of the overall performance of rip$^2$ALM. These results deepen our theoretical understanding of the family of proximal augmented Lagrangian methods and motivate their development for practical, large-scale structured application problems.

math.OC↗
arXiv2026

Scaling Spatial Intelligence with Multimodal Foundation Models

Despite remarkable progress, multimodal foundation models still exhibit surprising deficiencies in spatial intelligence. In this work, we explore scaling up multimodal foundation models to cultivate spatial intelligence within the SenseNova-SI family, built upon established multimodal foundations including visual understanding models (i.e., Qwen3-VL and InternVL3) and unified understanding and generation models (i.e., Bagel). We take a principled approach to constructing high-performing and robust spatial intelligence by systematically curating SenseNova-SI-8M: eight million diverse data samples under a rigorous taxonomy of spatial capabilities. SenseNova-SI demonstrates unprecedented performance across a broad range of spatial intelligence benchmarks: 68.8% on VSI-Bench, 43.3% on MMSI, 85.7% on MindCube, 54.7% on ViewSpatial, 47.7% on SITE, 63.9% on BLINK, 55.5% on 3DSR, and 72.0% on EmbSpatial, while maintaining strong general multimodal understanding (e.g., 84.9% on MMBench-En). More importantly, we analyze the impact of data scaling, discuss early signs of emergent generalization capabilities enabled by diverse data training, analyze the risk of overfitting and language shortcuts, present a preliminary study on spatial chain-of-thought reasoning, and validate the potential downstream application. All newly trained multimodal foundation models are publicly released.

cs.CV↗
arXiv2026

Phased DMD: Few-step Distribution Matching Distillation via Score Matching within Subintervals

Distribution Matching Distillation (DMD) distills score-based generative models into efficient one-step generators, without requiring a one-to-one correspondence with the sampling trajectories of their teachers. Yet, the limited capacity of one-step distilled models compromises generative diversity and degrades performance in complex generative tasks, e.g., generating intricate object motions in text-to-video task. Directly extending DMD to multi-step distillation increases memory usage and computational depth, leading to instability and reduced efficiency. While prior works propose stochastic gradient truncation as a potential solution, we observe that it substantially reduces the generative diversity in text-to-image generation and slows motion dynamics in video generation, reducing performance to the level of one-step models. To address these limitations, we propose Phased DMD, a multi-step distillation framework that bridges the idea of phase-wise distillation with Mixture-of-Experts (MoE), reducing learning difficulty while enhancing model capacity. Phased DMD incorporates two key ideas: progressive distribution matching and score matching within subintervals. First, our model divides the SNR range into subintervals, progressively refining the model to higher SNR levels, to better capture complex distributions. Next, to ensure accurate training within each subinterval, we derive rigorous mathematical formulations for the objective. We validate Phased DMD by distilling state-of-the-art image and video generation models, including Qwen-Image-20B and Wan2.2-28B. Experiments demonstrate that Phased DMD enhances motion dynamics, improves visual fidelity in video generation, and increases output diversity in image generation. Our code and models are available at https://x-niper.github.io/projects/Phased-DMD/.

cs.CV↗
arXiv2026

Identification and estimation of causal mechanisms in cluster-randomized trials with post-treatment confounding using Bayesian nonparametrics

Causal mediation analysis in cluster-randomized trials (CRTs) is essential for explaining how cluster-level interventions affect individual outcomes, yet it is complicated by interference, post-treatment confounding, and hierarchical covariate adjustment. We develop a Bayesian nonparametric framework that simultaneously accommodates interference and a post-treatment confounder that precedes the mediator. Identification is achieved through a multivariate Gaussian copula that replaces cross-world independence with a single dependence parameter, yielding a built-in sensitivity analysis to residual post-treatment confounding. For estimation, we introduce a nested common atoms enriched Dirichlet process (CA-EDP) prior that integrates the Common Atoms Model (CAM) to share information across clusters while capturing between- and within-cluster heterogeneity, and an Enriched Dirichlet Process (EDP) structure delivering robust covariate adjustment without impacting the outcome model. We provide formal theoretical support for our prior by deriving the model's key distributional properties, including its partially exchangeable partition structure, and by establishing convergence guarantees for the practical truncation-based posterior inference strategy. We demonstrate the performance of the proposed methods in simulations and provide further illustration through a reanalysis of a completed CRT.

stat.ME↗
arXiv2026

Cascade-Free Mandarin Visual Speech Recognition via Semantic-Guided Cross-Representation Alignment

Chinese mandarin visual speech recognition (VSR) is a task that has advanced in recent years, yet still lags behind the performance on non-tonal languages such as English. One primary challenge arises from the tonal nature of Mandarin, which limits the effectiveness of conventional sequence-to-sequence modeling approaches. To alleviate this issue, existing Chinese VSR systems commonly incorporate intermediate representations, most notably pinyin, within cascade architectures to enhance recognition accuracy. While beneficial, in these cascaded designs, the subsequent stage during inference depends on the output of the preceding stage, leading to error accumulation and increased inference latency. To address these limitations, we propose a cascade-free architecture based on multitask learning that jointly integrates multiple intermediate representations, including phoneme and viseme, to better exploit contextual information. The proposed semantic-guided local contrastive loss temporally aligns the features, enabling on-demand activation during inference, thereby providing a trade-off between inference efficiency and performance while mitigating error accumulation caused by projection and re-embedding. Experiments conducted on publicly available datasets demonstrate that our method achieves superior recognition performance.

cs.CV↗
arXiv2026

Fast and Effective Computation of Generalized Symmetric Matrix Factorization

In this paper, we study a nonconvex, nonsmooth, and non-Lipschitz generalized symmetric matrix factorization model that unifies a broad class of matrix factorization formulations arising in machine learning, image science, engineering, and related areas. We first establish two exactness properties. On the modeling side, we prove an exact penalty property showing that, under suitable conditions, the symmetry-inducing quadratic penalty enforces symmetry whenever the penalty parameter is sufficiently large but finite, thereby exactly recovering the associated symmetric formulation. On the algorithmic side, we introduce an auxiliary-variable splitting formulation and establish an exact relaxation relationship that rigorously links stationary points of the original objective function to those of a relaxed potential function. Building on these exactness properties, we propose an average-type nonmonotone alternating updating method (A-NAUM) based on the relaxed potential function. At each iteration, A-NAUM alternately updates the two factor blocks by (approximately) minimizing the potential function, while the auxiliary block is updated in closed form. To ensure the convergence and enhance practical performance, we further incorporate an average-type nonmonotone line search and show that it is well-defined under mild conditions. Moreover, based on the Kurdyka-Łojasiewicz property and its associated exponent, we establish global convergence of the entire sequence to a stationary point and derive convergence rate results. Finally, numerical experiments on real datasets demonstrate the efficiency of A-NAUM.

math.OC↗
arXiv2026

Bridging Semantic and Kinematic Conditions with Diffusion-based Discrete Motion Tokenizer

Prior motion generation largely follows two paradigms: continuous diffusion models that excel at kinematic control, and discrete token-based generators that are effective for semantic conditioning. To combine their strengths, we propose a three-stage framework comprising condition feature extraction (Perception), discrete token generation (Planning), and diffusion-based motion synthesis (Control). Central to this framework is MoTok, a diffusion-based discrete motion tokenizer that decouples semantic abstraction from fine-grained reconstruction by delegating motion recovery to a diffusion decoder, enabling compact single-layer tokens while preserving motion fidelity. For kinematic conditions, coarse constraints guide token generation during planning, while fine-grained constraints are enforced during control through diffusion-based optimization. This design prevents kinematic details from disrupting semantic token planning. On HumanML3D, our method significantly improves controllability and fidelity over MaskControl while using only one-sixth of the tokens, reducing trajectory error from 0.72 cm to 0.08 cm and FID from 0.083 to 0.029. Unlike prior methods that degrade under stronger kinematic constraints, ours improves fidelity, reducing FID from 0.033 to 0.014.

cs.CV↗
arXiv2026

A Delayed Radio Flare Traces Kinetic Energy Injection in the SMBHB Candidate SDSS~J143016.05+230344.4

SDSS~J143016.05+230344.4 ($z=0.08105$) has been proposed as a candidate pre-coalescence supermassive black hole binary and shows remarkable multiwavelength variability. Its radio evolution provides a direct probe of the compact emitting region and of the physical origin of the late-time activity. We aim to localize the variable radio emission, characterize its spectral evolution, and constrain whether the radio brightening is produced by a newly emerging compact component, external absorption, or dissipation in a structured circumnuclear environment. At all epochs, the radio emission is dominated by a single unresolved milliarcsecond core with $T_{\rm B} \gtrsim 10^{7}$ K, constraining the variable emission to $\lesssim 0.3$ pc. The broadband spectra require two synchrotron self-absorbed components: a persistent low-frequency component with $ν_{\rm p,steady} \approx 0.74$ GHz and $S_{\rm p,steady} \approx 1.22$ mJy, and a flare component whose turnover evolves from $(6.35 {\rm GHz}, 0.18 {\rm mJy})$ in 2022 February-May to $(8.61 {\rm GHz}, 0.38 {\rm mJy})$ in 2022 December, and then to $(5.83 {\rm GHz}, 0.25 {\rm mJy})$ in 2023 March-April. The flare contribution at 15 GHz reaches $\sim 80\%$ and matches the near-epoch VLBI recovery fraction, showing that the high-frequency brightening arises from a newly formed compact synchrotron component. A second brightening of the 15.2 GHz VLBI core is detected between 2023 September and 2024 February, while the source remains unresolved. Equipartition scalings imply characteristic radii of $\sim 5 \times 10^{-4}$ pc for the flare and $\sim 9 \times 10^{-3}$ pc for the steady component, and indicate a steep inner circumnuclear density profile, $n \propto R^{-1.7}$. The delayed radio flare is best explained by dissipation in an outflow or jet-base disturbance propagating through a structured circumnuclear medium.

astro-ph.HE↗
arXiv2026

ConsistCompose: Unified Multimodal Layout Control for Image Composition

Unified multimodal models that couple visual understanding with image generation have advanced rapidly, yet most systems still focus on visual grounding-aligning language with image regions-while their generative counterpart, linguistic-embedded layout-grounded generation (LELG) for layout-controllable multi-instance generation, remains underexplored and limits precise compositional control. We present ConsistCompose, a unified multimodal framework that embeds layout coordinates directly into language prompts, enabling layout-controlled multi-instance image generation from Interleaved Image-Text within a single generative interface. We further construct ConsistCompose3M, a 3.4M multi-instance generation dataset with layout and identity annotations (2.6M text-guided and 0.8M image-guided data pairs) that provides large-scale supervision for layout-conditioned generation. Within this framework, LELG is instantiated through instance-coordinate binding prompts and coordinate-aware classifier-free guidance, which translate linguistic layout cues into precise spatial control without task-specific branches. Experiments on COCO-Position and MS-Bench show that ConsistCompose substantially improves spatial accuracy over layout-controlled baselines while preserving identity fidelity and competitive general multimodal understanding, establishing a unified paradigm for layout-controllable multimodal image generation.

cs.CV↗
arXiv2026

Orbital-Selective Engineering of Strain-Tunable Chern Insulators in Momentum Space

Unlike conventional approaches where topological order is statically fixed post-synthesis, we demonstrate that a single external knob-strain-can independently modulate topological order and functional responses in the Tc-adsorbed penta-hexa silicene (Tc_PH-Si) monolayer, with both properties governed by a single microscopic mechanism: momentum-space orbital-selective engineering of Tc-dxz_Si-px hybridization. Combining first-principles calculations and tight-binding models, we show that biaxial strain drives a complete topological pathway: C=1 (0) to C=0 (-2) to C = -1 (-3 to -4) to C = 0 metallic state (-6). This is exemplified by two pivotal states: a topologically critical point yet functionally optimal state at -2 strain (C=0) hosting a direct bandgap (0.17 eV) and d11 = 8.34 pm_V, and a topologically nontrivial but equally optimal state at -4 strain (C = -1) with d11 = 11.01 pm_V-three times that of MoS2. Berry curvature analysis reveals that functionality arises from local orbital hybridization strength, while topology originates from its global phase distribution. This establishes a new paradigm for materials design, transforming static functional materials into dynamically tunable quantum platforms.

cond-mat.mtrl-sci↗
arXiv2026

Gluon Polarimetry with Energy-Energy Correlators

We propose a novel method to probe gluon linear polarization via energy correlations in hard scattering processes. This approach exploits the characteristic $\cos 2ϕ$ azimuthal modulation in single- and two-point energy correlations within jets initiated by polarized gluons. In contrast to conventional techniques that rely on $k_t$ resummation or intricate jet substructure observables, our method offers a theoretically robust and experimentally accessible avenue for gluon polarimetry. We perform an all-order analysis within the Ciafaloni-Catani-Fiorani-Marchesini (CCFM) formalism, incorporating coherent branching effects to achieve improved precision. Our predictions can be tested at current and future facilities, including the LHC, RHIC, HERA, and the EIC.

hep-ph↗
arXiv2026

Permutation invariant multi-scale full quantum neural network wavefunction

Solving the intricate quantum behavior of interacting particles is key to unlocking the mysteries of condensed matter, but capturing their complex correlations across different scales remains a monumental challenge. We introduce a neural network framework that overcomes this barrier by modeling the full quantum wavefunction of a system, including electrons, nuclei and muons, directly capturing the full quantum effects beyond the Born-Oppenheimer approximation. The neural network approximates joint wavefunction of different interacting particles with a rigorous handling of permutation invariance, enabling simultaneous treatment of nuclear quantum effects and electron-nucleus-muon couplings without explicit excited states. Validated on molecular systems, this approach offers a computationally feasible way to model full quantum phenomena in complex many-body systems, establishing a direct connection between fundamental particle properties and emergent material behavior.

physics.chem-ph↗
arXiv2026

Impact of Nuclear Reaction Rate Uncertainties on Type I X-ray Burst Nucleosynthesis: A Monte Carlo Study

Type I X-ray bursts are thermonuclear flashes on the surface of accreting neutron stars, involving hundreds of nuclei and thousands of reactions with larger uncertainties in reaction rate. To investigate the impact of nuclear reaction rate uncertainties on type I X-ray burst nucleosynthesis, comprehensive Monte Carlo simulations were performed with temperature-independent and -dependent variations in reaction rates using the REACLIB and STARLIB libraries, respectively. A total of 1,711 $(p, γ)$, $(p, α)$, $(α, p)$, and $(α, γ)$ reaction rates are varied simultaneously along with their inverse reactions via detailed balance. For the first time, it has been found that Monte Carlo sampling with larger perturbations to these reaction rates may lead to multi-peak abundance distributions for certain isotopes, such as $^{64}$Zn and $^{55}$Co. These multi-peak structures arise not only from coupled reactions but also from single reactions in some cases. Our studies also confirm previously identified key reactions and provide more robust lists that deserve priority consideration in future studies.

astro-ph.HE↗
arXiv2026

Finite-Sample Decision Instability in Threshold-Based Process Capability Approval

Process capability indices such as $C_{pk}$ are widely used in manufacturing quality control to support supplier qualification and product release decisions based on fixed acceptance thresholds (e.g., $C_{pk} \geq 1.33$). In practice, these decisions rely on sample-based estimates computed from moderate sample sizes ($n \approx$ 20-50), yet the stochastic nature of the estimator is often overlooked when interpreting threshold compliance. This study establishes a local asymptotic characterization of decision behavior when the true process capability lies near a fixed threshold. Under standard regularity conditions, if the true capability equals the threshold, the acceptance probability converges to 0.5 as sample size increases, implying that a fixed $C_{pk}$ gate embeds an inherent boundary decision risk even under ideal distributional assumptions. When the true capability deviates from the threshold by $O(n^{-1/2})$, the decision probability converges to a non-degenerate limit governed by a scaled signal-to-noise ratio. Monte Carlo simulations and an empirical study on 880 manufacturing dimensions demonstrate substantial resampling-based decision instability near the commonly used 1.33 criterion. These findings provide a probabilistic interpretation of threshold-based capability decisions and quantitative guidance for assessing boundary-induced release risk in engineering practice.

stat.AP↗