Abstract:We introduce VisualPRM, an advanced multimodal Process Reward Model (PRM) with 8B parameters, which improves the reasoning abilities of existing Multimodal Large Language Models (MLLMs) across different model scales and families with Best-of-N (BoN) evaluation strategies. Specifically, our model improves the reasoning performance of three types of MLLMs and four different model scales. Even when applied to the highly capable InternVL2.5-78B, it achieves a 5.9-point improvement across seven multimodal reasoning benchmarks. Experimental results show that our model exhibits superior performance compared to Outcome Reward Models and Self-Consistency during BoN evaluation. To facilitate the training of multimodal PRMs, we construct a multimodal process supervision dataset VisualPRM400K using an automated data pipeline. For the evaluation of multimodal PRMs, we propose VisualProcessBench, a benchmark with human-annotated step-wise correctness labels, to measure the abilities of PRMs to detect erroneous steps in multimodal reasoning tasks. We hope that our work can inspire more future research and contribute to the development of MLLMs. Our model, data, and benchmark are released in https://internvl.github.io/blog/2025-03-13-VisualPRM/.
Abstract:Existing motion generation methods based on mocap data are often limited by data quality and coverage. In this work, we propose a framework that generates diverse, physically feasible full-body human reaching and grasping motions using only brief walking mocap data. Base on the observation that walking data captures valuable movement patterns transferable across tasks and, on the other hand, the advanced kinematic methods can generate diverse grasping poses, which can then be interpolated into motions to serve as task-specific guidance. Our approach incorporates an active data generation strategy to maximize the utility of the generated motions, along with a local feature alignment mechanism that transfers natural movement patterns from walking data to enhance both the success rate and naturalness of the synthesized motions. By combining the fidelity and stability of natural walking with the flexibility and generalizability of task-specific generated data, our method demonstrates strong performance and robust adaptability in diverse scenes and with unseen objects.
Abstract:Parametric message passing (MP) is a promising technique that provides reliable marginal probability distributions for distributed cooperative positioning (DCP) based on factor graphs (FG), while maintaining minimal computational complexity. However, conventional parametric MP-based DCP methods may fail to converge in dense wireless networks due to numerous short loops on FG. Additionally, the use of inappropriate message approximation techniques can lead to increased sensitivity to initial values and significantly slower convergence rates. To address the challenging DCP problem modeled by a loopy FG, we propose an effective graph neural network enhanced fast convergent parametric MP (GNN--FCPMP) method. We first employ Chebyshev polynomials to approximate the nonlinear terms present in the FG-based spatio-temporal messages. This technique facilitates the derivation of globally precise, closed-form representations for each message transmitted across the FG. Then, the parametric representations of spatial messages are meticulously refined through data-driven graph neural networks (GNNs). Conclusively, by performing inference on the FG, we derive more accurate closed-form expressions for the a posteriori distributions of node positions. Numerical results substantiate the capability of GNN--FCPMP to significantly enhance positioning accuracy within wireless networks characterized by high-density loops and ensure rapid convergence.
Abstract:Current robot learning algorithms for acquiring novel skills often rely on demonstration datasets or environment interactions, resulting in high labor costs and potential safety risks. To address these challenges, this study proposes a skill-learning framework that enables robots to acquire novel skills from natural language instructions. The proposed pipeline leverages vision-language models to generate demonstration videos of novel skills, which are processed by an inverse dynamics model to extract actions from the unlabeled demonstrations. These actions are subsequently mapped to environmental contexts via imitation learning, enabling robots to learn new skills effectively. Experimental evaluations in the MetaWorld simulation environments demonstrate the pipeline's capability to generate high-fidelity and reliable demonstrations. Using the generated demonstrations, various skill learning algorithms achieve an accomplishment rate three times the original on novel tasks. These results highlight a novel approach to robot learning, offering a foundation for the intuitive and intelligent acquisition of novel robotic skills.
Abstract:Physical adversarial attacks in driving scenarios can expose critical vulnerabilities in visual perception models. However, developing such attacks remains challenging due to diverse real-world backgrounds and the requirement for maintaining visual naturality. Building upon this challenge, we reformulate physical adversarial attacks as a one-shot patch-generation problem. Our approach generates adversarial patches through a deep generative model that considers the specific scene context, enabling direct physical deployment in matching environments. The primary challenge lies in simultaneously achieving two objectives: generating adversarial patches that effectively mislead object detection systems while determining contextually appropriate placement within the scene. We propose MAGIC (Mastering Physical Adversarial Generation In Context), a novel framework powered by multi-modal LLM agents to address these challenges. MAGIC automatically understands scene context and orchestrates adversarial patch generation through the synergistic interaction of language and vision capabilities. MAGIC orchestrates three specialized LLM agents: The adv-patch generation agent (GAgent) masters the creation of deceptive patches through strategic prompt engineering for text-to-image models. The adv-patch deployment agent (DAgent) ensures contextual coherence by determining optimal placement strategies based on scene understanding. The self-examination agent (EAgent) completes this trilogy by providing critical oversight and iterative refinement of both processes. We validate our method on both digital and physical level, \ie, nuImage and manually captured real scenes, where both statistical and visual results prove that our MAGIC is powerful and effectively for attacking wide-used object detection systems.
Abstract:We introduce InternVL 2.5, an advanced multimodal large language model (MLLM) series that builds upon InternVL 2.0, maintaining its core model architecture while introducing significant enhancements in training and testing strategies as well as data quality. In this work, we delve into the relationship between model scaling and performance, systematically exploring the performance trends in vision encoders, language models, dataset sizes, and test-time configurations. Through extensive evaluations on a wide range of benchmarks, including multi-discipline reasoning, document understanding, multi-image / video understanding, real-world comprehension, multimodal hallucination detection, visual grounding, multilingual capabilities, and pure language processing, InternVL 2.5 exhibits competitive performance, rivaling leading commercial models such as GPT-4o and Claude-3.5-Sonnet. Notably, our model is the first open-source MLLMs to surpass 70% on the MMMU benchmark, achieving a 3.7-point improvement through Chain-of-Thought (CoT) reasoning and showcasing strong potential for test-time scaling. We hope this model contributes to the open-source community by setting new standards for developing and applying multimodal AI systems. HuggingFace demo see https://huggingface.co/spaces/OpenGVLab/InternVL
Abstract:Large vision-language models (LVLMs) have shown remarkable capabilities in interpreting visual content. While existing works demonstrate these models' vulnerability to deliberately placed adversarial texts, such texts are often easily identifiable as anomalous. In this paper, we present the first approach to generate scene-coherent typographic adversarial attacks that mislead advanced LVLMs while maintaining visual naturalness through the capability of the LLM-based agent. Our approach addresses three critical questions: what adversarial text to generate, where to place it within the scene, and how to integrate it seamlessly. We propose a training-free, multi-modal LLM-driven scene-coherent typographic adversarial planning (SceneTAP) that employs a three-stage process: scene understanding, adversarial planning, and seamless integration. The SceneTAP utilizes chain-of-thought reasoning to comprehend the scene, formulate effective adversarial text, strategically plan its placement, and provide detailed instructions for natural integration within the image. This is followed by a scene-coherent TextDiffuser that executes the attack using a local diffusion mechanism. We extend our method to real-world scenarios by printing and placing generated patches in physical environments, demonstrating its practical implications. Extensive experiments show that our scene-coherent adversarial text successfully misleads state-of-the-art LVLMs, including ChatGPT-4o, even after capturing new images of physical setups. Our evaluations demonstrate a significant increase in attack success rates while maintaining visual naturalness and contextual appropriateness. This work highlights vulnerabilities in current vision-language models to sophisticated, scene-coherent adversarial attacks and provides insights into potential defense mechanisms.
Abstract:Existing open-source multimodal large language models (MLLMs) generally follow a training process involving pre-training and supervised fine-tuning. However, these models suffer from distribution shifts, which limit their multimodal reasoning, particularly in the Chain-of-Thought (CoT) performance. To address this, we introduce a preference optimization (PO) process to enhance the multimodal reasoning capabilities of MLLMs. Specifically, (1) on the data side, we design an automated preference data construction pipeline to create MMPR, a high-quality, large-scale multimodal reasoning preference dataset. and (2) on the model side, we explore integrating PO with MLLMs, developing a simple yet effective method, termed Mixed Preference Optimization (MPO), which boosts multimodal CoT performance. Our approach demonstrates improved performance across multiple benchmarks, particularly in multimodal reasoning tasks. Notably, our model, InternVL2-8B-MPO, achieves an accuracy of 67.0 on MathVista, outperforming InternVL2-8B by 8.7 points and achieving performance comparable to the 10x larger InternVL2-76B. We hope this study could inspire further advancements in MLLMs. Code, data, and model shall be publicly released.
Abstract:Despite significant advancements in Multimodal Large Language Models (MLLMs) for understanding complex human intentions through cross-modal interactions, capturing intricate image details remains challenging. Previous methods integrating multiple vision encoders to enhance visual detail introduce redundancy and computational overhead. We observe that most MLLMs utilize only the last-layer feature map of the vision encoder for visual representation, neglecting the rich fine-grained information in shallow feature maps. To address this issue, we propose \modelname, a simple yet effective multi-layer feature fuser that efficiently integrates deep and shallow features from Vision Transformers (ViTs). Specifically, it leverages semantically aligned deep features as queries to dynamically extract missing details from shallow features, thus preserving semantic alignment while enriching the representation with fine-grained information. Applied to the LLaVA-1.5 model, \modelname~achieves significant improvements in visual representation and benchmark performance, providing a more flexible and lightweight solution compared to multi-encoder ensemble methods. The code and model have been released at https://github.com/yuecao0119/MMFuser.
Abstract:Attention mechanisms, particularly channel attention, have become highly influential in numerous computer vision tasks. Despite their effectiveness, many existing methods primarily focus on optimizing performance through complex attention modules applied at individual convolutional layers, often overlooking the synergistic interactions that can occur across multiple layers. In response to this gap, we introduce bridge attention, a novel approach designed to facilitate more effective integration and information flow between different convolutional layers. Our work extends the original bridge attention model (BAv1) by introducing an adaptive selection operator, which reduces information redundancy and optimizes the overall information exchange. This enhancement results in the development of BAv2, which achieves substantial performance improvements in the ImageNet classification task, obtaining Top-1 accuracies of 80.49% and 81.75% when using ResNet50 and ResNet101 as backbone networks, respectively. These results surpass the retrained baselines by 1.61% and 0.77%, respectively. Furthermore, BAv2 outperforms other existing channel attention techniques, such as the classical SENet101, exceeding its retrained performance by 0.52% Additionally, integrating BAv2 into advanced convolutional networks and vision transformers has led to significant gains in performance across a wide range of computer vision tasks, underscoring its broad applicability.