arXiv subjects

Chen Li

Publications and source records attributed to Chen Li.

At least 127 records · Page 7Linked to original sources

arXiv2026

Creating a biologically more accurate spider robot to study active vibration sensing

Orb-weaving spiders detect prey on a web using vibration sensors at leg joints. They often dynamically crouch their legs during prey sensing, likely an active sensing strategy. However, how leg crouching enhances sensing is poorly understood, because measuring system vibrations in behaving animals is difficult. We use robophysical modeling to study this problem. Our previous spider robot had only four legs, simplified leg morphology, and a shallow crouching range of motion. Here, we developed a new spider robot, with eight legs, each with four joints that better approximated spider leg morphology. Leg exoskeletons were 3-D printed and joint stiffness was tuned using integrated silicone molding with variable materials and geometry. Tendon-driven actuation allowed a motor in the body to crouch all eight legs deeply as spiders do, while accelerometers at leg joints record leg vibrations. Experiments showed that our new spider robot reproduced key vibration features observed in the previous robot while improving biological accuracy. Our new robot provides a biologically more accurate robophysical model for studying how leg behaviors modulate vibration sensing on a web.

cs.RO↗

arXiv2025

Why orb-weaving spiders use leg crouching behavior in vibration sensing of prey on a web: A physical mechanism from robophysical modeling

Orb-weaving spiders primarily sense leg vibrations to detect and locate prey caught on their wheel-shaped webs. Biological experiments and computational modeling elucidated the physics of how these spiders use long-timescale web-building behaviors, which occur before prey capture, to modulate vibration sensing of prey by controlling web geometry, materials, and tension distribution. By contrast, the physics of how spiders use short-timescale leg behaviors to modulate vibration sensing on a web during prey capture is less known. This is in part due to challenges in biological experiments (e.g., having little control over spider behavior, difficulty measuring the whole spider-web-prey system vibrations) and theoretical/computation modeling (e.g., close-form equations intractable for a complex web, high computation cost for simulating vibrations with behaving animals). Here, we use robophysical modeling as a complementary approach to address these challenges and study how dynamic leg crouching behavior common in orb-weaving spiders contributes to vibration sensing of prey on a web. We created a robophysical model consisting of a spider robot that can dynamically crouch its legs and sense its leg vibrations and a prey robot that can shake both on a horizontal physical wheel-shaped web. We discovered a physical mechanism that can allow the spider to use leg crouching to detect prey presence and distance. Our work demonstrated that robophysical modeling is a useful approach for discovering physical mechanisms of how spiders use short-time scale leg behaviors to enhance vibration sensing of objects on a web and providing new biological hypotheses.

physics.bio-ph↗

arXiv2025

V-Thinker: Interactive Thinking with Images

Empowering Large Multimodal Models (LMMs) to deeply integrate image interaction with long-horizon reasoning capabilities remains a long-standing challenge in this field. Recent advances in vision-centric reasoning explore a promising "Thinking with Images" paradigm for LMMs, marking a shift from image-assisted reasoning to image-interactive thinking. While this milestone enables models to focus on fine-grained image regions, progress remains constrained by limited visual tool spaces and task-specific workflow designs. To bridge this gap, we present V-Thinker, a general-purpose multimodal reasoning assistant that enables interactive, vision-centric thinking through end-to-end reinforcement learning. V-Thinker comprises two key components: (1) a Data Evolution Flywheel that automatically synthesizes, evolves, and verifies interactive reasoning datasets across three dimensions-diversity, quality, and difficulty; and (2) a Visual Progressive Training Curriculum that first aligns perception via point-level supervision, then integrates interactive reasoning through a two-stage reinforcement learning framework. Furthermore, we introduce VTBench, an expert-verified benchmark targeting vision-centric interactive reasoning tasks. Extensive experiments demonstrate that V-Thinker consistently outperforms strong LMM-based baselines in both general and interactive reasoning scenarios, providing valuable insights for advancing image-interactive reasoning applications.

cs.CV↗

arXiv2025

Video Generation with Stable Transparency via Shiftable RGB-A Distribution Learner

Generating RGB-A videos, which include alpha channels for transparency, has wide applications. However, current methods often suffer from low quality due to confusion between RGB and alpha. In this paper, we address this problem by learning shiftable RGB-A distributions. We adjust both the latent space and noise space, shifting the alpha distribution outward while preserving the RGB distribution, thereby enabling stable transparency generation without compromising RGB quality. Specifically, for the latent space, we propose a transparency-aware bidirectional diffusion loss during VAE training, which shifts the RGB-A distribution according to likelihood. For the noise space, we propose shifting the mean of diffusion noise sampling and applying a Gaussian ellipse mask to provide transparency guidance and controllability. Additionally, we construct a high-quality RGB-A video dataset. Compared to state-of-the-art methods, our model excels in visual quality, naturalness, transparency rendering, inference convenience, and controllability. The released model is available on our website: https://donghaotian123.github.io/Wan-Alpha/.

cs.CV↗

arXiv2025

Delving into the depths of NGC 3783 with XRISM IV. Mapping of the accretion flow with Fe K$α$ emission lines

Using XRISM/Resolve $439 \, \rm ks$ time-averaged spectra of the well-known Seyfert-1.5 active galactic nucleus (AGN) in NGC 3783, we investigate the nature of the Fe K$α$ emission line at 6.4 keV, the strongest and most common X-ray line observed in AGN. Even the narrow component of the line is resolved with evident Fe K$α_{1}$ (6.404 keV) and K$α_{2}$ (6.391 keV) contributions in a 2:1 flux ratio, fully consistent with a neutral gas with negligible bulk velocity. The narrow and intermediate-width components have a full-width at half maximum (FWHM) of 350 $\pm$ 50 km/s and $3510 \pm 470 \, \rm km/s$, respectively, suggesting that they arise in the outer disk/torus and/or BLR. We detect a $10\%$ excess flux around 4 $-$ 7 keV that is not well described by a symmetric Gaussian line, but is consistent with a relativistically broadened emission line. In this paper, we take the simplest approach to model the asymmetric line as a single emission line (assuming either neutral, He-like or H-like iron) convolved with a relativistic disk line model. As expected, the inferred inclination angle is highly sensitive to the assumed ionization state, and ranges between $i=17-44^{ \circ}$. This model also constrains the black hole spin via the extent of the red wing: the required gravitational redshift in the fitted disk-line profile disfavors a non-spinning (Schwarzschild) black hole. The derived inner radius is close to the radius of the innermost stable circular orbit $r_{\rm ISCO}$ and strongly correlated with the black hole spin. To better constrain the spin, we fix the inner radius at $r_{\rm ISCO}$ and derive a lower limit on the spin of $a \ge 0.29$ at the 3 $σ$ confidence level. A Compton shoulder is detected in our data as well as a $2-3 \, σ$ detection of the Cr K$α$ and Ni K$α$ lines.

astro-ph.HE↗

arXiv2025

Delving into the depths of NGC 3783 with XRISM III. Birth of an ultrafast outflow during a soft flare

The 2024 X-ray/UV observation campaign of NGC 3783, led by XRISM, revealed the launch of an ultrafast outflow (UFO) with a radial velocity of 0.19c (57000 km/s). This event is synchronized with the sharp decay, within less than half a day, of a prominent soft X-ray/UV flare. Accounting for the look-elsewhere effect, the XRISM Resolve data alone indicate a low probability of 2e-5 that this UFO detection is due to random chance. The UFO features narrow H-like and He-like Fe lines with a velocity dispersion of 1000 km/s, suggesting that it originates from a dense clump. Beyond this primary detection, there are hints of weaker outflow signatures throughout the rise and fall phases of the soft flare. Their velocities increase from 0.05c to 0.3c over approximately three days, and they may be associated with a larger stream in which the clump is embedded. The radiation pressure is insuffcient to drive the acceleration of this rapidly evolving outflow. The observed evolution of the outflow kinematics instead closely resembles that of solar coronal mass ejections, implying magnetic driving and, conceivably, reconnection near the accretion disk as the likely mechanisms behind both the UFO launch and the associated soft flare.

astro-ph.HE↗

arXiv2025

VOST-SGG: VLM-Aided One-Stage Spatio-Temporal Scene Graph Generation

Spatio-temporal scene graph generation (ST-SGG) aims to model objects and their evolving relationships across video frames, enabling interpretable representations for downstream reasoning tasks such as video captioning and visual question answering. Despite recent advancements in DETR-style single-stage ST-SGG models, they still suffer from several key limitations. First, while these models rely on attention-based learnable queries as a core component, these learnable queries are semantically uninformed and instance-agnostically initialized. Second, these models rely exclusively on unimodal visual features for predicate classification. To address these challenges, we propose VOST-SGG, a VLM-aided one-stage ST-SGG framework that integrates the common sense reasoning capabilities of vision-language models (VLMs) into the ST-SGG pipeline. First, we introduce the dual-source query initialization strategy that disentangles what to attend to from where to attend, enabling semantically grounded what-where reasoning. Furthermore, we propose a multi-modal feature bank that fuses visual, textual, and spatial cues derived from VLMs for improved predicate classification. Extensive experiments on the Action Genome dataset demonstrate that our approach achieves state-of-the-art performance, validating the effectiveness of integrating VLM-aided semantic priors and multi-modal features for ST-SGG. We will release the code at https://github.com/LUNAProject22/VOST.

cs.CV↗

arXiv2025

VACoT: Rethinking Visual Data Augmentation with VLMs

While visual data augmentation remains a cornerstone for training robust vision models, it has received limited attention in visual language models (VLMs), which predominantly rely on large-scale real data acquisition or synthetic diversity. Consequently, they may struggle with basic perception tasks that conventional models handle reliably. Given the substantial cost of pre-training and fine-tuning VLMs, continue training on augmented data yields limited and diminishing returns. In this paper, we present Visual Augmentation Chain-of-Thought (VACoT), a framework that dynamically invokes image augmentations during model inference. By incorporating post-hoc transformations such as denoising, VACoT substantially improves robustness on challenging and out-of-distribution inputs, especially in OCR-related adversarial scenarios. Distinct from prior approaches limited to local cropping, VACoT integrates a structured collection of general visual augmentations, broadening the query image views while reducing training complexity and computational overhead with efficient agentic reinforcement learning. We propose a conditional reward scheme that encourages necessary augmentation while penalizing verbose responses, ensuring concise and effective reasoning in perception tasks. We demonstrate the superiority of VACoT with extensive experiments on 13 perception benchmarks and further introduce AdvOCR to highlight the generalization benefits of post-hoc visual augmentations in adversarial scenarios.

cs.CV↗

arXiv2025

DiverseAR: Boosting Diversity in Bitwise Autoregressive Image Generation

In this paper, we investigate the underexplored challenge of sample diversity in autoregressive (AR) generative models with bitwise visual tokenizers. We first analyze the factors that limit diversity in bitwise AR models and identify two key issues: (1) the binary classification nature of bitwise modeling, which restricts the prediction space, and (2) the overly sharp logits distribution, which causes sampling collapse and reduces diversity. Building on these insights, we propose DiverseAR, a principled and effective method that enhances image diversity without sacrificing visual quality. Specifically, we introduce an adaptive logits distribution scaling mechanism that dynamically adjusts the sharpness of the binary output distribution during sampling, resulting in smoother predictions and greater diversity. To mitigate potential fidelity loss caused by distribution smoothing, we further develop an energy-based generation path search algorithm that avoids sampling low-confidence tokens, thereby preserving high visual quality. Extensive experiments demonstrate that DiverseAR substantially improves sample diversity in bitwise autoregressive image generation.

cs.CV↗

arXiv2025

UnicEdit-10M: A Dataset and Benchmark Breaking the Scale-Quality Barrier via Unified Verification for Reasoning-Enriched Edits

With the rapid advances of powerful multimodal models such as GPT-4o, Nano Banana, and Seedream 4.0 in Image Editing, the performance gap between closed-source and open-source models is widening, primarily due to the scarcity of large-scale, high-quality training data and comprehensive benchmarks capable of diagnosing model weaknesses across diverse editing behaviors. Existing data construction methods face a scale-quality trade-off: human annotations are high-quality but not scalable, while automated pipelines suffer from error propagation and noise. To address this, we introduce a lightweight data pipeline that replaces multi-toolchains with an end-to-end model and a unified post-verification stage. For scalable quality control, we train a 7B dual-task expert model, \textbf{Qwen-Verify}, for efficient failure detection and instruction recaptioning. This pipeline yields \textbf{UnicEdit-10M}, a 10M-scale dataset spanning diverse basic and complex editing tasks. We also propose \textbf{UnicBench}, a general benchmark that extends beyond basic edits to explicitly assess spatial and knowledge-driven reasoning. To enable fine-grained diagnosis, we introduce novel metrics, including \textit{Non-edit Consistency} and \textit{Reasoning Accuracy}. Our analysis of mainstream models on UnicBench reveals their limitations and provides clear directions for future research.

cs.CV↗

arXiv2025

HMR3D: Hierarchical Multimodal Representation for 3D Scene Understanding with Large Vision-Language Model

Recent advances in large vision-language models (VLMs) have shown significant promise for 3D scene understanding. Existing VLM-based approaches typically align 3D scene features with the VLM's embedding space. However, this implicit alignment often yields suboptimal performance due to the scarcity of 3D data and the inherent complexity of spatial relationships in 3D environments. To address these limitations, we propose a novel hierarchical multimodal representation for 3D scene reasoning that explicitly aligns with VLMs at the input space by leveraging both multi-view images and text descriptions. The text descriptions capture spatial relationships by referencing the 3D coordinates of detected objects, while the multi-view images include a top-down perspective and four directional views (forward, left, right, and backward), ensuring comprehensive scene coverage. Additionally, we introduce a hierarchical feature representation that aggregates patch-level image features into view-level and scene-level representations, enabling the model to reason over both local and global scene context. Experimental results on both situated 3D Q&A and general 3D Q&A benchmarks demonstrate the effectiveness of our approach.

cs.CV↗

arXiv2025

HPSU: A Benchmark for Human-Level Perception in Real-World Spoken Speech Understanding

Recent advances in Speech Large Language Models (Speech LLMs) have led to great progress in speech understanding tasks such as Automatic Speech Recognition (ASR) and Speech Emotion Recognition (SER). However, whether these models can achieve human-level auditory perception, particularly in terms of their ability to comprehend latent intentions and implicit emotions in real-world spoken language, remains underexplored. To this end, we introduce the Human-level Perception in Spoken Speech Understanding (HPSU), a new benchmark for fully evaluating the human-level perceptual and understanding capabilities of Speech LLMs. HPSU comprises over 20,000 expert-validated spoken language understanding samples in English and Chinese. It establishes a comprehensive evaluation framework by encompassing a spectrum of tasks, ranging from basic speaker attribute recognition to complex inference of latent intentions and implicit emotions. To address the issues of data scarcity and high cost of manual annotation in real-world scenarios, we developed a semi-automatic annotation process. This process fuses audio, textual, and visual information to enable precise speech understanding and labeling, thus enhancing both annotation efficiency and quality. We systematically evaluate various open-source and proprietary Speech LLMs. The results demonstrate that even top-performing models still fall considerably short of human capabilities in understanding genuine spoken interactions. Consequently, HPSU will be useful for guiding the development of Speech LLMs toward human-level perception and cognition.

cs.SD↗

arXiv2025

FlowSteer: Guiding Few-Step Image Synthesis with Authentic Trajectories

With the success of flow matching in visual generation, sampling efficiency remains a critical bottleneck for its practical application. Among flow models' accelerating methods, ReFlow has been somehow overlooked although it has theoretical consistency with flow matching. This is primarily due to its suboptimal performance in practical scenarios compared to consistency distillation and score distillation. In this work, we investigate this issue within the ReFlow framework and propose FlowSteer, a method unlocks the potential of ReFlow-based distillation by guiding the student along teacher's authentic generation trajectories. We first identify that Piecewised ReFlow's performance is hampered by a critical distribution mismatch during the training and propose Online Trajectory Alignment(OTA) to resolve it. Then, we introduce a adversarial distillation objective applied directly on the ODE trajectory, improving the student's adherence to the teacher's generation trajectory. Furthermore, we find and fix a previously undiscovered flaw in the widely-used FlowMatchEulerDiscreteScheduler that largely degrades few-step inference quality. Our experiment result on SD3 demonstrates our method's efficacy.

cs.CV↗

arXiv2025

ADF-LoRA: Alternating Low-Rank Aggregation for Decentralized Federated Fine-Tuning

This paper revisits alternating low-rank updates for federated fine-tuning and examines their behavior in decentralized federated learning (DFL). While alternating the LoRA matrices has been shown to stabilize aggregation in centralized FL, extending this mechanism to decentralized, peer-to-peer communication introduces new challenges due to phase-state mismatch and block-wise divergence across clients. We introduce ADF-LoRA, which synchronizes the update of only one low-rank matrix per round and mixes both matrices to maintain more consistent parameter states under decentralized propagation. This design preserves the cross-term suppression effect of alternating updates while improving stability in serverless topologies. We provide a convergence analysis under standard smoothness assumptions and evaluate ADF-LoRA on multiple GLUE tasks. Experiments show that ADF-LoRA achieves faster and smoother convergence and delivers the highest average accuracy across tasks, outperforming existing LoRA variants in decentralized FL by a consistent margin.

cs.LG↗

arXiv2025

MCTSr-Zero: Self-Reflective Psychological Counseling Dialogues Generation via Principles and Adaptive Exploration

The integration of Monte Carlo Tree Search (MCTS) with Large Language Models (LLMs) has demonstrated significant success in structured, problem-oriented tasks. However, applying these methods to open-ended dialogues, such as those in psychological counseling, presents unique challenges. Unlike tasks with objective correctness, success in therapeutic conversations depends on subjective factors like empathetic engagement, ethical adherence, and alignment with human preferences, for which strict "correctness" criteria are ill-defined. Existing result-oriented MCTS approaches can therefore produce misaligned responses. To address this, we introduce MCTSr-Zero, an MCTS framework designed for open-ended, human-centric dialogues. Its core innovation is "domain alignment", which shifts the MCTS search objective from predefined end-states towards conversational trajectories that conform to target domain principles (e.g., empathy in counseling). Furthermore, MCTSr-Zero incorporates "Regeneration" and "Meta-Prompt Adaptation" mechanisms to substantially broaden exploration by allowing the MCTS to consider fundamentally different initial dialogue strategies. We evaluate MCTSr-Zero in psychological counseling by generating multi-turn dialogue data, which is used to fine-tune an LLM, PsyLLM. We also introduce PsyEval, a benchmark for assessing multi-turn psychological counseling dialogues. Experiments demonstrate that PsyLLM achieves state-of-the-art performance on PsyEval and other relevant metrics, validating MCTSr-Zero's effectiveness in generating high-quality, principle-aligned conversational data for human-centric domains and addressing the LLM challenge of consistently adhering to complex psychological standards.

cs.CL↗

arXiv2025

ARC-Chapter: Structuring Hour-Long Videos into Navigable Chapters and Hierarchical Summaries

The proliferation of hour-long videos (e.g., lectures, podcasts, documentaries) has intensified demand for efficient content structuring. However, existing approaches are constrained by small-scale training with annotations that are typical short and coarse, restricting generalization to nuanced transitions in long videos. We introduce ARC-Chapter, the first large-scale video chaptering model trained on over million-level long video chapters, featuring bilingual, temporally grounded, and hierarchical chapter annotations. To achieve this goal, we curated a bilingual English-Chinese chapter dataset via a structured pipeline that unifies ASR transcripts, scene texts, visual captions into multi-level annotations, from short title to long summaries. We demonstrate clear performance improvements with data scaling, both in data volume and label intensity. Moreover, we design a new evaluation metric termed GRACE, which incorporates many-to-one segment overlaps and semantic similarity, better reflecting real-world chaptering flexibility. Extensive experiments demonstrate that ARC-Chapter establishes a new state-of-the-art by a significant margin, outperforming the previous best by 14.0% in F1 score and 11.3% in SODA score. Moreover, ARC-Chapter shows excellent transferability, improving the state-of-the-art on downstream tasks like dense video captioning on YouCook2.

cs.CV↗

arXiv2025

Beyond Born-Oppenheimer Time-Dependent Density Functional Theory

We formulate a time-dependent density functional theory for the coupled dynamics of electrons and nuclei that goes beyond the Born-Oppenheimer (BO) approximation. We prove that the time-dependent marginal nuclear probability density $|χ({\bdu R},t)|^2$, the conditional electronic density $n_{\bdu R}(\br,t)$, and the current density $\bm J_{\bdu R}(\br,t)$ are sufficient to uniquely determine the full time-evolving electron-nuclear wave function, and thus the dynamics of all observables. Moreover, we propose a time-dependent Kohn-Sham scheme which reproduces the exact conditional electronic density and current density and the exact N-body nuclear density. The remaining task is to look for functional approximations for the Kohn-Sham exchange-correlation scalar and vector potentials. Using a model driven proton transfer system, we numerically demonstrate that the adiabatic extension of a beyond-BO ground state functional captures the dominant nonadiabatic effects in the regime of slow driving.

physics.chem-ph↗

arXiv2025

FabasedVC: Enhancing Voice Conversion with Text Modality Fusion and Phoneme-Level SSL Features

In voice conversion (VC), it is crucial to preserve complete semantic information while accurately modeling the target speaker's timbre and prosody. This paper proposes FabasedVC to achieve VC with enhanced similarity in timbre, prosody, and duration to the target speaker, as well as improved content integrity. It is an end-to-end VITS-based VC system that integrates relevant textual modality information, phoneme-level self-supervised learning (SSL) features, and a duration predictor. Specifically, we employ a text feature encoder to encode attributes such as text, phonemes, tones and BERT features. We then process the frame-level SSL features into phoneme-level features using two methods: average pooling and attention mechanism based on each phoneme's duration. Moreover, a duration predictor is incorporated to better align the speech rate and prosody of the target speaker. Experimental results demonstrate that our method outperforms competing systems in terms of naturalness, similarity, and content integrity.

cs.SD↗