arXiv subjects

Changsheng Xu

Publications and source records attributed to Changsheng Xu.

At least 37 records · Page 2Linked to original sources

arXiv2026

General Covariant Action Modeling: Constructing Generalized Manifolds via Spatio-Temporal Decoupling

Achieving robust generalization from limited data is a central challenge in embodied intelligence. Prevailing methods fail by regressing absolute coordinates, which violates the principle of general covariance. Fundamentally, this conflates the intrinsic task geometry with rigid execution patterns, binding policies to specific motion styles and fixed speeds. To resolve this, we propose the Generalized Action Manifold (GAM) framework that enforces general covariance through structural disentanglement. Specifically, GAM realizes the manifold by enforcing invariance across two orthogonal dimensions: (1) Temporal Invariance, utilizing an Arc-Length Parameterizer to orthogonalize the spatial path geometry from temporal dynamics, ensuring robustness to velocity variations; (2) Geometric Invariance, where a Schema-Affine-Factorization mechanism maps trajectories to canonical ``world lines'' in a pose-normalized coordinate frame. This distinguishes invariant geometric schemas from affine modulations, ensuring spatial generalizability. By integrating GAM within a structured Vision-Language-Action (VLA) architecture, we enable sparse demonstrations to densely populate a continuous, valid action manifold. Empirical results demonstrate that GAM enables superior transfer and robustness capabilities, outperforming geometry-agnostic baselines.

cs.CV↗

arXiv2026

LAST: Bridging Vision-Language and Action Manifolds via Gromov-Wasserstein Alignment

We take a Gromov-Wasserstein perspective on Vision-Language-Action (VLA) learning, where the goal is to make the relational geometry of action representations compatible with the semantic geometry of VL embeddings. However, this alignment is non-trivial due to the mathematical heterogeneity between the domains: the semantic space of vision-language is topologically linear and isotropic, whereas the physical manifold of robotic action is non-Euclidean and anisotropic. Their disjoint metric structures render direct regression ill-posed. To resolve this incompatibility, we introduce LAST (Lie-algebraic Action Space Tokenizer), which reconstructs the action space to establish local metric compatibility with the VL modality via a two-stage transformation: (1) Global Topological Linearization: linearizing the action manifold via Lie-algebraic mapping, converting trajectories into a fixed-length, physically additive representation. (2) Local Metric Discretization: hierarchically discretizing the representation into schemas and whitened residuals, yielding approximately isotropic local charts that are statistically aligned with the semantic metric. By resolving the structural mismatch at both global and local levels, LAST enables VLA models with superior convergence and generalizability.

cs.CV↗

arXiv2026

Replacing Parameters with Preferences: Federated Alignment of Heterogeneous Vision-Language Models

Vision-Language Models (VLMs) have broad potential in privacy-sensitive domains such as healthcare and finance, yet strict data-sharing constraints render centralized training infeasible. Federated Learning mitigates this issue by enabling decentralized training, but practical deployments face challenges due to client heterogeneity in computational resources, application requirements, and model architectures. Under extreme model and data heterogeneity, replacing parameter aggregation with preference-based collaboration offers a more suitable interface, as it eliminates the need for direct parameter or data exchange. Motivated by this, we propose MoR, a federated alignment framework that combines GRPO with Mixture-of-Rewards for heterogeneous VLMs. In MoR, each client locally trains a reward model from local preference annotations, capturing specific evaluation signals without exposing raw data. To combine these heterogeneous supervision signals, MoR introduces a Mixture-of-Rewards mechanism with learned routing, which adaptively fuses client reward models according to the input and alignment objective. The server then optimizes a base VLM using GRPO with a KL penalty to a reference model, enabling preference alignment without requiring client models to share architectures or parameters. Experiments on diverse public vision-language benchmarks demonstrate that MoR consistently outperforms federated alignment baselines in generalization and cross-client adaptability. Our approach provides a scalable solution for privacy-preserving alignment of heterogeneous VLMs under federated settings.

cs.AI↗

arXiv2026

A Step Toward Federated Pretraining of Multimodal Large Language Models

The rapid evolution of Multimodal Large Language Models (MLLMs) is bottlenecked by the saturation of high-quality public data, while vast amounts of diverse multimodal data remain inaccessible in privacy-sensitive silos. Federated Learning (FL) offers a promising solution to unlock these distributed resources, but existing research focuses predominantly on fine-tuning, leaving the foundational pre-training phase largely unexplored. In this paper, we formally introduce the Federated MLLM Alignment (Fed-MA) task, a lightweight pre-training paradigm that freezes the vision encoder and LLM while collaboratively training the cross-modal projector. We identify two critical challenges in this setting: (i) parameter interference in aggregating local projectors; and (ii) gradient oscillations in one-pass collaborative SGD. To address these challenges, we propose Fed-CMP, a pioneering framework for federated MLLM pre-training. Fed-CMP employs Canonical Reliability-Aware Aggregation, which constructs a canonical space to decompose client projectors into a shared alignment basis and client-specific coefficients, then performs reliability-weighted fusion to suppress parameter interference. Furthermore, Fed-CMP introduces Orthogonality-Preserved Momentum, which applies momentum to the shared alignment basis via orthogonal projection, accumulating historical optimization directions while preserving geometric structure. We construct four federated pre-training scenarios based on public datasets, and extensive experiments validate that Fed-CMP significantly outperforms existing baselines.

cs.LG↗

arXiv2026

Complementary Text-Guided Attention for Zero-Shot Adversarial Robustness

Due to the impressive zero-shot capabilities, pre-trained vision-language models (e.g., CLIP), have attracted widespread attention and adoption across various domains. Nonetheless, CLIP has been observed to be susceptible to adversarial examples. Through experimental analysis, we have observed a phenomenon wherein adversarial perturbations induce shifts in text-guided attention. Building upon this observation, we propose a simple yet effective strategy: Text-Guided Attention for Zero-Shot Robustness (TGA-ZSR). This framework incorporates two components: Local Attention Refinement Module and Global Attention Constraint Module. Our goal is to maintain the generalization of the CLIP model and enhance its adversarial robustness. Additionally, the Global Attention Constraint Module acquires text-guided attention from both the target and original models using clean examples. Its objective is to maintain model performance on clean samples while enhancing overall robustness. However, we observe that the method occasionally focuses on irrelevant or spurious features, which can lead to suboptimal performance and undermine its robustness in certain scenarios. To overcome this limitation, we further propose a novel approach called Complementary Text-Guided Attention (Comp-TGA). This method integrates two types of foreground attention: attention guided by the class prompt and reversed attention driven by the non-class prompt. These complementary attention mechanisms allow the model to capture a more comprehensive and accurate representation of the foreground. The experiments validate that TGA-ZSR and Comp-TGA yield 9.58% and 11.95% improvements respectively, in zero-shot robust accuracy over the current state-of-the-art techniques across 16 datasets.

cs.CV↗

arXiv2026

Replacing Parameters with Preferences: Federated Alignment of Heterogeneous Vision-Language Models

VLMs have broad potential in privacy-sensitive domains such as healthcare and finance, yet strict data-sharing constraints render centralized training infeasible. FL mitigates this issue by enabling decentralized training, but practical deployments face challenges due to client heterogeneity in computational resources, application requirements, and model architectures. We argue that while replacing data with model parameters characterizes the present of FL, replacing parameters with preferences represents a more scalable and privacy-preserving future. Motivated by this perspective, we propose MoR, a federated alignment framework based on GRPO with Mixture-of-Rewards for heterogeneous VLMs. MoR initializes a visual foundation model as a KL-regularized reference, while each client locally trains a reward model from local preference annotations, capturing specific evaluation signals without exposing raw data. To reconcile heterogeneous rewards, we introduce a routing-based fusion mechanism that adaptively aggregates client reward signals. Finally, the server performs GRPO with this mixed reward to optimize the base VLM. Experiments on three public VQA benchmarks demonstrate that MoR consistently outperforms federated alignment baselines in generalization, robustness, and cross-client adaptability. Our approach provides a scalable solution for privacy-preserving alignment of heterogeneous VLMs under federated settings.

cs.AI↗

arXiv2026

A Survey on Cross-Modal Interaction Between Music and Multimodal Data

Multimodal learning has driven innovation across various industries, particularly in the field of music. By enabling more intuitive interaction experiences and enhancing immersion, it not only lowers the entry barriers to the music but also increases its overall appeal. This survey aims to provide a comprehensive review of multimodal tasks related to music, outlining how music contributes to multimodal learning and offering insights for researchers seeking to expand the boundaries of computational music. Unlike text and images, which are often semantically or visually intuitive, music primarily interacts with humans through auditory perception, making its data representation inherently less intuitive. Therefore, this paper first introduces the representations of music and provides an overview of music datasets. Subsequently, we categorize cross-modal interactions between music and multimodal data into three types: music-driven cross-modal interactions, music-oriented cross-modal interactions, and bidirectional music cross-modal interactions. For each category, we systematically trace the development of relevant sub-tasks, analyze existing limitations, and discuss emerging trends. Furthermore, we provide a comprehensive summary of datasets and evaluation metrics used in multimodal tasks related to music, offering benchmark references for future research. Finally, we discuss the current challenges in cross-modal interactions involving music and propose potential directions for future research.

cs.MM↗

arXiv2025

Dual Cluster Contrastive learning for Object Re-Identification

Recently, cluster contrastive learning has been proven effective for object ReID by computing the contrastive loss between the individual features and the cluster memory. However, existing methods that use the individual features to momentum update the cluster memory will fluctuate over the training examples, especially for the outlier samples. Unlike the individual-based updating mechanism, the centroid-based updating mechanism that applies the mean feature of each cluster to update the cluster memory can reduce the impact of individual samples. Therefore, we formulate the individual-based updating and centroid-based updating mechanisms in a unified cluster contrastive framework, named Dual Cluster Contrastive framework (DCC), which maintains two types of memory banks: individual and centroid cluster memory banks. Significantly, the individual cluster memory considers just one individual at a time to take a single step for updating. The centroid cluster memory applies the mean feature of each cluster to update the corresponding cluster memory. During optimization, besides the vallina contrastive loss of each memory, a cross-view consistency constraint is applied to exchange the benefits of two memories for generating a discriminative description for the object ReID. Note that DCC can be easily applied for unsupervised or supervised object ReID by using ground-truth labels or the generated pseudo-labels. Extensive experiments on three benchmarks, \emph{e.g.,} Market-1501, MSMT17, and VeRi-776, under \textbf{supervised Object ReID} and \textbf{unsupervised Object ReID} demonstrate the superiority of the proposed DCC.

cs.CV↗

arXiv2025

SoMe: A Realistic Benchmark for LLM-based Social Media Agents

Intelligent agents powered by large language models (LLMs) have recently demonstrated impressive capabilities and gained increasing popularity on social media platforms. While LLM agents are reshaping the ecology of social media, there exists a current gap in conducting a comprehensive evaluation of their ability to comprehend media content, understand user behaviors, and make intricate decisions. To address this challenge, we introduce SoMe, a pioneering benchmark designed to evaluate social media agents equipped with various agent tools for accessing and analyzing social media data. SoMe comprises a diverse collection of 8 social media agent tasks, 9,164,284 posts, 6,591 user profiles, and 25,686 reports from various social media platforms and external websites, with 17,869 meticulously annotated task queries. Compared with the existing datasets and benchmarks for social media tasks, SoMe is the first to provide a versatile and realistic platform for LLM-based social media agents to handle diverse social media tasks. By extensive quantitative and qualitative analysis, we provide the first overview insight into the performance of mainstream agentic LLMs in realistic social media environments and identify several limitations. Our evaluation reveals that both the current closed-source and open-source LLMs cannot handle social media agent tasks satisfactorily. SoMe provides a challenging yet meaningful testbed for future social media agents. Our code and data are available at https://github.com/LivXue/SoMe

cs.SI↗

arXiv2025

DMC$^3$: Dual-Modal Counterfactual Contrastive Construction for Egocentric Video Question Answering

Egocentric Video Question Answering (Egocentric VideoQA) plays an important role in egocentric video understanding, which refers to answering questions based on first-person videos. Although existing methods have made progress through the paradigm of pre-training and fine-tuning, they ignore the unique challenges posed by the first-person perspective, such as understanding multiple events and recognizing hand-object interactions. To deal with these challenges, we propose a Dual-Modal Counterfactual Contrastive Construction (DMC$^3$) framework, which contains an egocentric videoqa baseline, a counterfactual sample construction module and a counterfactual sample-involved contrastive optimization. Specifically, We first develop a counterfactual sample construction module to generate positive and negative samples for textual and visual modalities through event description paraphrasing and core interaction mining, respectively. Then, We feed these samples together with the original samples into the baseline. Finally, in the counterfactual sample-involved contrastive optimization module, we apply contrastive loss to minimize the distance between the original sample features and the positive sample features, while maximizing the distance from the negative samples. Experiments show that our method achieve 52.51\% and 46.04\% on the \textit{normal} and \textit{indirect} splits of EgoTaskQA, and 13.2\% on QAEGO4D, both reaching the state-of-the-art performance.

cs.CV↗

arXiv2025

Modality-Collaborative Low-Rank Decomposers for Few-Shot Video Domain Adaptation

In this paper, we study the challenging task of Few-Shot Video Domain Adaptation (FSVDA). The multimodal nature of videos introduces unique challenges, necessitating the simultaneous consideration of both domain alignment and modality collaboration in a few-shot scenario, which is ignored in previous literature. We observe that, under the influence of domain shift, the generalization performance on the target domain of each individual modality, as well as that of fused multimodal features, is constrained. Because each modality is comprised of coupled features with multiple components that exhibit different domain shifts. This variability increases the complexity of domain adaptation, thereby reducing the effectiveness of multimodal feature integration. To address these challenges, we introduce a novel framework of Modality-Collaborative LowRank Decomposers (MC-LRD) to decompose modality-unique and modality-shared features with different domain shift levels from each modality that are more friendly for domain alignment. The MC-LRD comprises multiple decomposers for each modality and Multimodal Decomposition Routers (MDR). Each decomposer has progressively shared parameters across different modalities. The MDR is leveraged to selectively activate the decomposers to produce modality-unique and modality-shared features. To ensure efficient decomposition, we apply orthogonal decorrelation constraints separately to decomposers and subrouters, enhancing their diversity. Furthermore, we propose a cross-domain activation consistency loss to guarantee that target and source samples of the same category exhibit consistent activation preferences of the decomposers, thereby facilitating domain alignment. Extensive experimental results on three public benchmarks demonstrate that our model achieves significant improvements over existing methods.

cs.CV↗

arXiv2025

Look Before You Leap: A GUI-Critic-R1 Model for Pre-Operative Error Diagnosis in GUI Automation

In recent years, Multimodal Large Language Models (MLLMs) have been extensively utilized for multimodal reasoning tasks, including Graphical User Interface (GUI) automation. Unlike general offline multimodal tasks, GUI automation is executed in online interactive environments, necessitating step-by-step decision-making based on real-time status of the environment. This task has a lower tolerance for decision-making errors at each step, as any mistakes may cumulatively disrupt the process and potentially lead to irreversible outcomes like deletions or payments. To address these issues, we introduce a pre-operative critic mechanism that provides effective feedback prior to the actual execution, by reasoning about the potential outcome and correctness of actions. Specifically, we propose a Suggestion-aware Gradient Relative Policy Optimization (S-GRPO) strategy to construct our pre-operative critic model GUI-Critic-R1, incorporating a novel suggestion reward to enhance the reliability of the model's feedback. Furthermore, we develop a reasoning-bootstrapping based data collection pipeline to create a GUI-Critic-Train and a GUI-Critic-Test, filling existing gaps in GUI critic data. Static experiments on the GUI-Critic-Test across both mobile and web domains reveal that our GUI-Critic-R1 offers significant advantages in critic accuracy compared to current MLLMs. Dynamic evaluation on GUI automation benchmark further highlights the effectiveness and superiority of our model, as evidenced by improved success rates and operational efficiency.

cs.AI↗

arXiv2025

SVBench: A Benchmark with Temporal Multi-Turn Dialogues for Streaming Video Understanding

Despite the significant advancements of Large Vision-Language Models (LVLMs) on established benchmarks, there remains a notable gap in suitable evaluation regarding their applicability in the emerging domain of long-context streaming video understanding. Current benchmarks for video understanding typically emphasize isolated single-instance text inputs and fail to evaluate the capacity to sustain temporal reasoning throughout the entire duration of video streams. To address these limitations, we introduce SVBench, a pioneering benchmark with temporal multi-turn question-answering chains specifically designed to thoroughly assess the capabilities of streaming video understanding of current LVLMs. We design a semi-automated annotation pipeline to obtain 49,979 Question-Answer (QA) pairs of 1,353 streaming videos, which includes generating QA chains that represent a series of consecutive multi-turn dialogues over video segments and constructing temporal linkages between successive QA chains. Our experimental results, obtained from 14 models in dialogue and streaming evaluations, reveal that while the closed-source GPT-4o outperforms others, most open-source LVLMs struggle with long-context streaming video understanding. We also construct a StreamingChat model, which significantly outperforms open-source LVLMs on our SVBench and achieves comparable performance on diverse vision-language benchmarks. We expect SVBench to advance the research of streaming video understanding by providing a comprehensive and in-depth analysis of current LVLMs. Our benchmark and model can be accessed at https://github.com/sotayang/SVBench.

cs.CV↗

arXiv2025

LiveStar: Live Streaming Assistant for Real-World Online Video Understanding

Despite significant progress in Video Large Language Models (Video-LLMs) for offline video understanding, existing online Video-LLMs typically struggle to simultaneously process continuous frame-by-frame inputs and determine optimal response timing, often compromising real-time responsiveness and narrative coherence. To address these limitations, we introduce LiveStar, a pioneering live streaming assistant that achieves always-on proactive responses through adaptive streaming decoding. Specifically, LiveStar incorporates: (1) a training strategy enabling incremental video-language alignment for variable-length video streams, preserving temporal consistency across dynamically evolving frame sequences; (2) a response-silence decoding framework that determines optimal proactive response timing via a single forward pass verification; (3) memory-aware acceleration via peak-end memory compression for online inference on 10+ minute videos, combined with streaming key-value cache to achieve 1.53x faster inference. We also construct an OmniStar dataset, a comprehensive dataset for training and benchmarking that encompasses 15 diverse real-world scenarios and 5 evaluation tasks for online video understanding. Extensive experiments across three benchmarks demonstrate LiveStar's state-of-the-art performance, achieving an average 19.5% improvement in semantic correctness with 18.1% reduced timing difference compared to existing online Video-LLMs, while improving FPS by 12.0% across all five OmniStar tasks. Our model and dataset can be accessed at https://github.com/yzy-bupt/LiveStar.

cs.CV↗

arXiv2025

StreamingCoT: A Dataset for Temporal Dynamics and Multimodal Chain-of-Thought Reasoning in Streaming VideoQA

The rapid growth of streaming video applications demands multimodal models with enhanced capabilities for temporal dynamics understanding and complex reasoning. However, current Video Question Answering (VideoQA) datasets suffer from two critical limitations: 1) Static annotation mechanisms fail to capture the evolving nature of answers in temporal video streams, and 2) The absence of explicit reasoning process annotations restricts model interpretability and logical deduction capabilities. To address these challenges, We introduce StreamingCoT, the first dataset explicitly designed for temporally evolving reasoning in streaming VideoQA and multimodal Chain-of-Thought (CoT) tasks. Our framework first establishes a dynamic hierarchical annotation architecture that generates per-second dense descriptions and constructs temporally-dependent semantic segments through similarity fusion, paired with question-answer sets constrained by temporal evolution patterns. We further propose an explicit reasoning chain generation paradigm that extracts spatiotemporal objects via keyframe semantic alignment, derives object state transition-based reasoning paths using large language models, and ensures logical coherence through human-verified validation. This dataset establishes a foundation for advancing research in streaming video understanding, complex temporal reasoning, and multimodal inference. Our StreamingCoT and its construction toolkit can be accessed at https://github.com/Fleeting-hyh/StreamingCoT.

cs.CV↗

arXiv2025

OmniSAT: Compact Action Token, Faster Auto Regression

Existing Vision-Language-Action (VLA) models can be broadly categorized into diffusion-based and auto-regressive (AR) approaches: diffusion models capture continuous action distributions but rely on computationally heavy iterative denoising. In contrast, AR models enable efficient optimization and flexible sequence construction, making them better suited for large-scale pretraining. To further improve AR efficiency, particularly when action chunks induce extended and high-dimensional sequences, prior work applies entropy-guided and token-frequency techniques to shorten the sequence length. However, such compression struggled with \textit{poor reconstruction or inefficient compression}. Motivated by this, we introduce an Omni Swift Action Tokenizer, which learns a compact, transferable action representation. Specifically, we first normalize value ranges and temporal horizons to obtain a consistent representation with B-Spline encoding. Then, we apply multi-stage residual quantization to the position, rotation, and gripper subspaces, producing compressed discrete tokens with coarse-to-fine granularity for each part. After pre-training on the large-scale dataset Droid, the resulting discrete tokenization shortens the training sequence by 6.8$\times$, and lowers the target entropy. To further explore the potential of OmniSAT, we develop a cross-embodiment learning strategy that builds on the unified action-pattern space and jointly leverages robot and human demonstrations. It enables scalable auxiliary supervision from heterogeneous egocentric videos. Across diverse real-robot and simulation experiments, OmniSAT encompasses higher compression while preserving reconstruction quality, enabling faster AR training convergence and model performance.

cs.CV↗

arXiv2025

Short-video Propagation Influence Rating: A New Real-world Dataset and A New Large Graph Model

Short-video platforms have gained immense popularity, captivating the interest of millions, if not billions, of users globally. Recently, researchers have highlighted the significance of analyzing the propagation of short-videos, which typically involves discovering commercial values, public opinions, user behaviors, etc. This paper proposes a new Short-video Propagation Influence Rating (SPIR) task and aims to promote SPIR from both the dataset and method perspectives. First, we propose a new Cross-platform Short-Video (XS-Video) dataset, which aims to provide a large-scale and real-world short-video propagation network across various platforms to facilitate the research on short-video propagation. Our XS-Video dataset includes 117,720 videos, 381,926 samples, and 535 topics across 5 biggest Chinese platforms, annotated with the propagation influence from level 0 to 9. To the best of our knowledge, this is the first large-scale short-video dataset that contains cross-platform data or provides all of the views, likes, shares, collects, fans, comments, and comment content. Second, we propose a Large Graph Model (LGM) named NetGPT, based on a novel three-stage training mechanism, to bridge heterogeneous graph-structured data with the powerful reasoning ability and knowledge of Large Language Models (LLMs). Our NetGPT can comprehend and analyze the short-video propagation graph, enabling it to predict the long-term propagation influence of short-videos. Comprehensive experimental results evaluated by both classification and regression metrics on our XS-Video dataset indicate the superiority of our method for SPIR.

cs.CV↗

arXiv2025

LumiSculpt: Enabling Consistent Portrait Lighting in Video Generation

Lighting plays a pivotal role in ensuring the naturalness and aesthetic quality of video generation. However, the impact of lighting is deeply coupled with other factors of videos, e.g., objects and scenes. Thus, it remains challenging to disentangle and model coherent lighting conditions independently, limiting the flexibility to control lighting in video generation. In this paper, inspired by the established controllable T2I models, we propose LumiSculpt, which enables precise and consistent lighting control in T2V generation models. LumiSculpt equips the video generation with new interactive capabilities, allowing the input of reference image sequences with customized lighting conditions. Furthermore, the core learnable plug-and-play module of LumiSculpt facilitates direct control over the intensity, position and trajectory of an assumed light source in video diffusion models. To effectively train LumiSculpt and address the issue of insufficient lighting data, we construct LumiHuman, a new lightweight and flexible dataset for portrait lighting of images and videos. Experimental results demonstrate that LumiSculpt achieves precise and high-quality lighting control in video generation. The analysis demonstrates the flexibility of LumiHuman.

cs.CV↗