Visual Language Models (VLMs) have demonstrated impressive capabilities in visual grounding tasks. However, their effectiveness in the medical domain, particularly for abnormality detection and localization within medical images, remains underexplored. A major challenge is the complex and abstract nature of medical terminology, which makes it difficult to directly associate pathological anomaly terms with their corresponding visual features. In this work, we introduce a novel approach to enhance VLM performance in medical abnormality detection and localization by leveraging decomposed medical knowledge. Instead of directly prompting models to recognize specific abnormalities, we focus on breaking down medical concepts into fundamental attributes and common visual patterns. This strategy promotes a stronger alignment between textual descriptions and visual features, improving both the recognition and localization of abnormalities in medical images.We evaluate our method on the 0.23B Florence-2 base model and demonstrate that it achieves comparable performance in abnormality grounding to significantly larger 7B LLaVA-based medical VLMs, despite being trained on only 1.5% of the data used for such models. Experimental results also demonstrate the effectiveness of our approach in both known and previously unseen abnormalities, suggesting its strong generalization capabilities.
Embodied intelligence requires agents to interact with 3D environments in real time based on language instructions. A foundational task in this domain is ego-centric 3D visual grounding. However, the point clouds rendered from RGB-D images retain a large amount of redundant background data and inherent noise, both of which can interfere with the manifold structure of the target regions. Existing point cloud enhancement methods often require a tedious process to improve the manifold, which is not suitable for real-time tasks. We propose Proxy Transformation suitable for multimodal task to efficiently improve the point cloud manifold. Our method first leverages Deformable Point Clustering to identify the point cloud sub-manifolds in target regions. Then, we propose a Proxy Attention module that utilizes multimodal proxies to guide point cloud transformation. Built upon Proxy Attention, we design a submanifold transformation generation module where textual information globally guides translation vectors for different submanifolds, optimizing relative spatial relationships of target regions. Simultaneously, image information guides linear transformations within each submanifold, refining the local point cloud manifold of target regions. Extensive experiments demonstrate that Proxy Transformation significantly outperforms all existing methods, achieving an impressive improvement of 7.49% on easy targets and 4.60% on hard targets, while reducing the computational overhead of attention blocks by 40.6%. These results establish a new SOTA in ego-centric 3D visual grounding, showcasing the effectiveness and robustness of our approach.
Recent advancements in software engineering (SWE) agents have largely followed a $\textit{tool-based paradigm}$, where agents interact with hand-engineered tool APIs to perform specific tasks. While effective for specialized tasks, these methods fundamentally lack generalization, as they require predefined tools for each task and do not scale across programming languages and domains. We introduce $\texttt{Programming with Pixels}$ (PwP), an agent environment that unifies software development tasks by enabling $\textit{computer-use agents}$-agents that operate directly within an IDE through visual perception, typing, and clicking, rather than relying on predefined tool APIs. To systematically evaluate these agents, we propose $\texttt{PwP-Bench}$, a benchmark that unifies existing SWE benchmarks spanning tasks across multiple programming languages, modalities, and domains under a task-agnostic state and action space. Our experiments demonstrate that general-purpose computer-use agents can approach or even surpass specialized tool-based agents on a variety of SWE tasks without the need for hand-engineered tools. However, our analysis shows that current models suffer from limited visual grounding and fail to exploit many IDE tools that could simplify their tasks. When agents can directly access IDE tools, without visual interaction, they show significant performance improvements, highlighting the untapped potential of leveraging built-in IDE capabilities. Our results establish PwP as a scalable testbed for building and evaluating the next wave of software engineering agents. We release code and data at https://programmingwithpixels.com
Visual grounding aims to ground an image region through natural language, which heavily relies on cross-modal alignment. Most existing methods transfer visual/linguistic knowledge separately by fully fine-tuning uni-modal pre-trained models, followed by a simple stack of visual-language transformers for multimodal fusion. However, these approaches not only limit adequate interaction between visual and linguistic contexts, but also incur significant computational costs. Therefore, to address these issues, we explore a step-wise multimodal fusion and adaption framework, namely SwimVG. Specifically, SwimVG proposes step-wise multimodal prompts (Swip) and cross-modal interactive adapters (CIA) for visual grounding, replacing the cumbersome transformer stacks for multimodal fusion. Swip can improve {the} alignment between the vision and language representations step by step, in a token-level fusion manner. In addition, weight-level CIA further promotes multimodal fusion by cross-modal interaction. Swip and CIA are both parameter-efficient paradigms, and they fuse the cross-modal features from shallow to deep layers gradually. Experimental results on four widely-used benchmarks demonstrate that SwimVG achieves remarkable abilities and considerable benefits in terms of efficiency. Our code is available at https://github.com/liuting20/SwimVG.
Current image captioning systems lack the ability to link descriptive text to specific visual elements, making their outputs difficult to verify. While recent approaches offer some grounding capabilities, they cannot track object identities across multiple references or ground both actions and objects simultaneously. We propose a novel ID-based grounding system that enables consistent object reference tracking and action-object linking, and present GroundCap, a dataset containing 52,016 images from 77 movies, with 344 human-annotated and 52,016 automatically generated captions. Each caption is grounded on detected objects (132 classes) and actions (51 classes) using a tag system that maintains object identity while linking actions to the corresponding objects. Our approach features persistent object IDs for reference tracking, explicit action-object linking, and segmentation of background elements through K-means clustering. We propose gMETEOR, a metric combining caption quality with grounding accuracy, and establish baseline performance by fine-tuning Pixtral-12B. Human evaluation demonstrates our approach's effectiveness in producing verifiable descriptions with coherent object references.
In this paper, we propose an efficient multi-level convolution architecture for 3D visual grounding. Conventional methods are difficult to meet the requirements of real-time inference due to the two-stage or point-based architecture. Inspired by the success of multi-level fully sparse convolutional architecture in 3D object detection, we aim to build a new 3D visual grounding framework following this technical route. However, as in 3D visual grounding task the 3D scene representation should be deeply interacted with text features, sparse convolution-based architecture is inefficient for this interaction due to the large amount of voxel features. To this end, we propose text-guided pruning (TGP) and completion-based addition (CBA) to deeply fuse 3D scene representation and text features in an efficient way by gradual region pruning and target completion. Specifically, TGP iteratively sparsifies the 3D scene representation and thus efficiently interacts the voxel features with text features by cross-attention. To mitigate the affect of pruning on delicate geometric information, CBA adaptively fixes the over-pruned region by voxel completion with negligible computational overhead. Compared with previous single-stage methods, our method achieves top inference speed and surpasses previous fastest method by 100\% FPS. Our method also achieves state-of-the-art accuracy even compared with two-stage methods, with $+1.13$ lead of [email omitted] on ScanRefer, and $+2.6$ and $+3.2$ leads on NR3D and SR3D respectively. The code is available at \href{https://github.com/GWxuan/TSP3D}{https://github.com/GWxuan/TSP3D}.
A robust and efficient traffic monitoring system is essential for smart cities and Intelligent Transportation Systems (ITS), using sensors and cameras to track vehicle movements, optimize traffic flow, reduce congestion, enhance road safety, and enable real-time adaptive traffic control. Traffic monitoring models must comprehensively understand dynamic urban conditions and provide an intuitive user interface for effective management. This research leverages the LLaVA visual grounding multimodal large language model (LLM) for traffic monitoring tasks on the real-time Quanser Interactive Lab simulation platform, covering scenarios like intersections, congestion, and collisions. Cameras placed at multiple urban locations collect real-time images from the simulation, which are fed into the LLaVA model with queries for analysis. An instance segmentation model integrated into the cameras highlights key elements such as vehicles and pedestrians, enhancing training and throughput. The system achieves 84.3% accuracy in recognizing vehicle locations and 76.4% in determining steering direction, outperforming traditional models.
In this work, we propose a modular approach for the Vision-Language Navigation (VLN) task by decomposing the problem into four sub-modules that use state-of-the-art Large Language Models (LLMs) and Vision-Language Models (VLMs) in a zero-shot setting. Given navigation instruction in natural language, we first prompt LLM to extract the landmarks and the order in which they are visited. Assuming the known model of the environment, we retrieve the top-k locations of the last landmark and generate $k$ path hypotheses from the starting location to the last landmark using the shortest path algorithm on the topological map of the environment. Each path hypothesis is represented by a sequence of panoramas. We then use dynamic programming to compute the alignment score between the sequence of panoramas and the sequence of landmark names, which match scores obtained from VLM. Finally, we compute the nDTW metric between the hypothesis that yields the highest alignment score to evaluate the path fidelity. We demonstrate superior performance compared to other approaches that use joint semantic maps like VLMaps on the complex R2R-Habitat instruction dataset and quantify in detail the effect of visual grounding on navigation performance.
3D visual grounding (3DVG) is challenging because of the requirement of understanding on visual information, language and spatial relationships. While supervised approaches have achieved superior performance, they are constrained by the scarcity and high cost of 3D vision-language datasets. On the other hand, LLM/VLM based agents are proposed for 3DVG, eliminating the need for training data. However, these methods incur prohibitive time and token costs during inference. To address the challenges, we introduce a novel training-free symbolic framework for 3D visual grounding, namely Evolvable Symbolic Visual Grounder, that offers significantly reduced inference costs compared to previous agent-based methods while maintaining comparable performance. EaSe uses LLM generated codes to compute on spatial relationships. EaSe also implements an automatic pipeline to evaluate and optimize the quality of these codes and integrate VLMs to assist in the grounding process. Experimental results demonstrate that EaSe achieves 52.9% accuracy on Nr3D dataset and 49.2% [email omitted] on ScanRefer, which is top-tier among training-free methods. Moreover, it substantially reduces the inference time and cost, offering a balanced trade-off between performance and efficiency. Codes are available at https://github.com/OpenRobotLab/EaSe.
Vision-language model (VLM) fine-tuning for application-specific visual grounding based on natural language instructions has become one of the most popular approaches for learning-enabled autonomous systems. However, such fine-tuning relies heavily on high-quality datasets to achieve successful performance in various downstream tasks. Additionally, VLMs often encounter limitations due to insufficient and imbalanced fine-tuning data. To address these issues, we propose a new generalizable framework to improve VLM fine-tuning by integrating it with a reinforcement learning (RL) agent. Our method utilizes the RL agent to manipulate objects within an indoor setting to create synthetic data for fine-tuning to address certain vulnerabilities of the VLM. Specifically, we use the performance of the VLM to provide feedback to the RL agent to generate informative data that efficiently fine-tune the VLM over the targeted task (e.g. spatial reasoning). The key contribution of this work is developing a framework where the RL agent serves as an informative data sampling tool and assists the VLM in order to enhance performance and address task-specific vulnerabilities. By targeting the data sampling process to address the weaknesses of the VLM, we can effectively train a more context-aware model. In addition, generating synthetic data allows us to have precise control over each scene and generate granular ground truth captions. Our results show that the proposed data generation approach improves the spatial reasoning performance of VLMs, which demonstrates the benefits of using RL-guided data generation in vision-language tasks.
Referring video object segmentation (RVOS) aims to segment target objects throughout a video based on a text description. Despite notable progress in recent years, current RVOS models remain struggle to handle complicated object descriptions due to their limited video-language understanding. To address this limitation, we present \textbf{ReferDINO}, an end-to-end RVOS model that inherits strong vision-language understanding from the pretrained visual grounding foundation models, and is further endowed with effective temporal understanding and object segmentation capabilities. In ReferDINO, we contribute three technical innovations for effectively adapting the foundation models to RVOS: 1) an object-consistent temporal enhancer that capitalizes on the pretrained object-text representations to enhance temporal understanding and object consistency; 2) a grounding-guided deformable mask decoder that integrates text and grounding conditions to generate accurate object masks; 3) a confidence-aware query pruning strategy that significantly improves the object decoding efficiency without compromising performance. We conduct extensive experiments on five public RVOS benchmarks to demonstrate that our proposed ReferDINO outperforms state-of-the-art methods significantly. Project page: \url{https://isee-laboratory.github.io/ReferDINO}
Large Vision Language Models (LVLMs) have demonstrated remarkable capabilities in understanding and describing visual content, achieving state-of-the-art performance across various vision-language tasks. However, these models often generate descriptions containing objects or details that are absent in the input image, a phenomenon commonly known as hallucination. Our work investigates the key reasons behind this issue by analyzing the pattern of self-attention in transformer layers. We find that hallucinations often arise from the progressive weakening of attention weight to visual tokens in the deeper layers of the LLM. Some previous works naively boost the attention of all visual tokens to mitigate this issue, resulting in suboptimal hallucination reduction. To address this, we identify two critical sets of visual tokens that facilitate the transfer of visual information from the vision encoder to the LLM. Local tokens encode grounded information about objects present in an image, while summary tokens capture the overall aggregated representation of the image. Importantly, these two sets of tokens require different levels of weight enhancement. To this end, we propose \textbf{PAINT} (\textbf{P}aying \textbf{A}ttention to \textbf{IN}formed \textbf{T}okens), a plug-and-play framework that intervenes in the self-attention mechanism of the LLM, selectively boosting the attention weights of local and summary tokens with experimentally learned margins. Evaluation on the MSCOCO image captioning dataset demonstrate that our approach reduces hallucination rates by up to 62.3\% compared to baseline models while maintaining accuracy. Code is available at \href{https://github.com/hasanar1f/PAINT}{https://github.com/hasanar1f/PAINT}
The increasing availability of traffic videos functioning on a 24/7/365 time scale has the great potential of increasing the spatio-temporal coverage of traffic accidents, which will help improve traffic safety. However, analyzing footage from hundreds, if not thousands, of traffic cameras in a 24/7/365 working protocol remains an extremely challenging task, as current vision-based approaches primarily focus on extracting raw information, such as vehicle trajectories or individual object detection, but require laborious post-processing to derive actionable insights. We propose SeeUnsafe, a new framework that integrates Multimodal Large Language Model (MLLM) agents to transform video-based traffic accident analysis from a traditional extraction-then-explanation workflow to a more interactive, conversational approach. This shift significantly enhances processing throughput by automating complex tasks like video classification and visual grounding, while improving adaptability by enabling seamless adjustments to diverse traffic scenarios and user-defined queries. Our framework employs a severity-based aggregation strategy to handle videos of various lengths and a novel multimodal prompt to generate structured responses for review and evaluation and enable fine-grained visual grounding. We introduce IMS (Information Matching Score), a new MLLM-based metric for aligning structured responses with ground truth. We conduct extensive experiments on the Toyota Woven Traffic Safety dataset, demonstrating that SeeUnsafe effectively performs accident-aware video classification and visual grounding by leveraging off-the-shelf MLLMs. Source code will be available at \url{https://github.com/ai4ce/SeeUnsafe}.
Object Referring Analysis (ORA), commonly known as referring expression comprehension, requires the identification and localization of specific objects in an image based on natural descriptions. Unlike generic object detection, ORA requires both accurate language understanding and precise visual localization, making it inherently more complex. Although recent pre-trained large visual grounding detectors have achieved significant progress, they heavily rely on extensively labeled data and time-consuming learning. To address these, we introduce a novel, training-free framework for zero-shot ORA, termed FLORA (Formal Language for Object Referring and Analysis). FLORA harnesses the inherent reasoning capabilities of large language models (LLMs) and integrates a formal language model - a logical framework that regulates language within structured, rule-based descriptions - to provide effective zero-shot ORA. More specifically, our formal language model (FLM) enables an effective, logic-driven interpretation of object descriptions without necessitating any training processes. Built upon FLM-regulated LLM outputs, we further devise a Bayesian inference framework and employ appropriate off-the-shelf interpretive models to finalize the reasoning, delivering favorable robustness against LLM hallucinations and compelling ORA performance in a training-free manner. In practice, our FLORA boosts the zero-shot performance of existing pretrained grounding detectors by up to around 45%. Our comprehensive evaluation across different challenging datasets also confirms that FLORA consistently surpasses current state-of-the-art zero-shot methods in both detection and segmentation tasks associated with zero-shot ORA. We believe our probabilistic parsing and reasoning of the LLM outputs elevate the reliability and interpretability of zero-shot ORA. We shall release codes upon publication.
3D visual grounding (3DVG), which aims to correlate a natural language description with the target object within a 3D scene, is a significant yet challenging task. Despite recent advancements in this domain, existing approaches commonly encounter a shortage: a limited amount and diversity of text3D pairs available for training. Moreover, they fall short in effectively leveraging different contextual clues (e.g., rich spatial relations within the 3D visual space) for grounding. To address these limitations, we propose AugRefer, a novel approach for advancing 3D visual grounding. AugRefer introduces cross-modal augmentation designed to extensively generate diverse text-3D pairs by placing objects into 3D scenes and creating accurate and semantically rich descriptions using foundation models. Notably, the resulting pairs can be utilized by any existing 3DVG methods for enriching their training data. Additionally, AugRefer presents a language-spatial adaptive decoder that effectively adapts the potential referring objects based on the language description and various 3D spatial relations. Extensive experiments on three benchmark datasets clearly validate the effectiveness of AugRefer.
The multimodal language models (MLMs) based on generative pre-trained Transformer are considered powerful candidates for unifying various domains and tasks. MLMs developed for remote sensing (RS) have demonstrated outstanding performance in multiple tasks, such as visual question answering and visual grounding. In addition to visual grounding that detects specific objects corresponded to given instruction, aerial detection, which detects all objects of multiple categories, is also a valuable and challenging task for RS foundation models. However, aerial detection has not been explored by existing RS MLMs because the autoregressive prediction mechanism of MLMs differs significantly from the detection outputs. In this paper, we present a simple baseline for applying MLMs to aerial detection for the first time, named LMMRotate. Specifically, we first introduce a normalization method to transform detection outputs into textual outputs to be compatible with the MLM framework. Then, we propose a evaluation method, which ensures a fair comparison between MLMs and conventional object detection models. We construct the baseline by fine-tuning open-source general-purpose MLMs and achieve impressive detection performance comparable to conventional detector. We hope that this baseline will serve as a reference for future MLM development, enabling more comprehensive capabilities for understanding RS images. Code is available at https://github.com/Li-Qingyun/mllm-mmrotate.
Current multimodal large language models (MLLMs) often underperform on mathematical problem-solving tasks that require fine-grained visual understanding. The limitation is largely attributable to inadequate perception of geometric primitives during image-level contrastive pre-training (e.g., CLIP). While recent efforts to improve math MLLMs have focused on scaling up mathematical visual instruction datasets and employing stronger LLM backbones, they often overlook persistent errors in visual recognition. In this paper, we systematically evaluate the visual grounding capabilities of state-of-the-art MLLMs and reveal a significant negative correlation between visual grounding accuracy and problem-solving performance, underscoring the critical role of fine-grained visual understanding. Notably, advanced models like GPT-4o exhibit a 70% error rate when identifying geometric entities, highlighting that this remains a key bottleneck in visual mathematical reasoning. To address this, we propose a novel approach, SVE-Math (Selective Vision-Enhanced Mathematical MLLM), featuring a geometric-grounded vision encoder and a feature router that dynamically adjusts the contribution of hierarchical visual feature maps. Our model recognizes accurate visual primitives and generates precise visual prompts tailored to the language model's reasoning needs. In experiments, SVE-Math-Qwen2.5-7B outperforms other 7B models by 15% on MathVerse and is compatible with GPT-4V on MathVista. Despite being trained on smaller datasets, SVE-Math-7B achieves competitive performance on GeoQA, rivaling models trained on significantly larger datasets. Our findings emphasize the importance of incorporating fine-grained visual understanding into MLLMs and provide a promising direction for future research.
Multi-modal large language models (MLLMs) have achieved remarkable success in image- and region-level remote sensing (RS) image understanding tasks, such as image captioning, visual question answering, and visual grounding. However, existing RS MLLMs lack the pixel-level dialogue capability, which involves responding to user instructions with segmentation masks for specific instances. In this paper, we propose GeoPix, a RS MLLM that extends image understanding capabilities to the pixel level. This is achieved by equipping the MLLM with a mask predictor, which transforms visual features from the vision encoder into masks conditioned on the LLM's segmentation token embeddings. To facilitate the segmentation of multi-scale objects in RS imagery, a class-wise learnable memory module is integrated into the mask predictor to capture and store class-wise geo-context at the instance level across the entire dataset. In addition, to address the absence of large-scale datasets for training pixel-level RS MLLMs, we construct the GeoPixInstruct dataset, comprising 65,463 images and 140,412 instances, with each instance annotated with text descriptions, bounding boxes, and masks. Furthermore, we develop a two-stage training strategy to balance the distinct requirements of text generation and masks prediction in multi-modal multi-task optimization. Extensive experiments verify the effectiveness and superiority of GeoPix in pixel-level segmentation tasks, while also maintaining competitive performance in image- and region-level benchmarks.
Multi-task visual grounding involves the simultaneous execution of localization and segmentation in images based on textual expressions. The majority of advanced methods predominantly focus on transformer-based multimodal fusion, aiming to extract robust multimodal representations. However, ambiguity between referring expression comprehension (REC) and referring image segmentation (RIS) is error-prone, leading to inconsistencies between multi-task predictions. Besides, insufficient multimodal understanding directly contributes to biased target perception. To overcome these challenges, we propose a Coarse-to-fine Consistency Constraints Visual Grounding architecture ($\text{C}^3\text{VG}$), which integrates implicit and explicit modeling approaches within a two-stage framework. Initially, query and pixel decoders are employed to generate preliminary detection and segmentation outputs, a process referred to as the Rough Semantic Perception (RSP) stage. These coarse predictions are subsequently refined through the proposed Mask-guided Interaction Module (MIM) and a novel explicit bidirectional consistency constraint loss to ensure consistent representations across tasks, which we term the Refined Consistency Interaction (RCI) stage. Furthermore, to address the challenge of insufficient multimodal understanding, we leverage pre-trained models based on visual-linguistic fusion representations. Empirical evaluations on the RefCOCO, RefCOCO+, and RefCOCOg datasets demonstrate the efficacy and soundness of $\text{C}^3\text{VG}$, which significantly outperforms state-of-the-art REC and RIS methods by a substantial margin. Code and model will be available at \url{https://github.com/Dmmm1997/C3VG}.
Chain-of-Thought (CoT) reasoning is widely used to enhance the mathematical reasoning capabilities of large language models (LLMs). The introduction of process supervision for CoT trajectories has sparked discussions on improving test-time scaling, thereby unlocking the System 2-style thinking capabilities of these models. However, in multimodal mathematical reasoning, the scarcity of high-quality CoT training data has hindered existing models from achieving both deliberate reasoning and fine-grained verification. In this work, we propose a novel framework that introduces System 2-style thinking to multimodal mathematical reasoning. We introduce a three-module CoT data synthesis process that integrates CoT distillation, trajectory-format rewriting, and format unification. This process generates MMathCoT-1M, a high-quality CoT reasoning instruction fine-tuning dataset. Furthermore, we implement a dual-view trajectory labeling automation that targets both visual grounding fidelity and deductive chain validity, resulting in the DualMath-1.1M dataset. The URSA-8B model, trained on MMathCoT-1M, achieves new state-of-the-art (SOTA) performance among similarly sized multimodal LLMs on six popular reasoning benchmarks. Training URSA-8B further on the DualMath-1.1M dataset yields URSA-RM-8B, a verifier that enhances URSA-8B's test-time performance and surpasses strong closed-source multimodal MLLMs like GPT-4o. The model weights, training data, and code have been open-sourced: https://github.com/URSA-MATH/URSA-MATH.