Referring video object segmentation aims to segment and track a target object in a video using a natural language prompt. Existing methods typically fuse visual and textual features in a highly entangled manner, processing multi-modal information together to generate per-frame masks. However, this approach often struggles with ambiguous target identification, particularly in scenes with multiple similar objects, and fails to ensure consistent mask propagation across frames. To address these limitations, we introduce FindTrack, a novel decoupled framework that separates target identification from mask propagation. FindTrack first adaptively selects a key frame by balancing segmentation confidence and vision-text alignment, establishing a robust reference for the target object. This reference is then utilized by a dedicated propagation module to track and segment the object across the entire video. By decoupling these processes, FindTrack effectively reduces ambiguities in target association and enhances segmentation consistency. We demonstrate that FindTrack outperforms existing methods on public benchmarks.
Reinforcement Fine-Tuning (RFT) in Large Reasoning Models like OpenAI o1 learns from feedback on its answers, which is especially useful in applications when fine-tuning data is scarce. Recent open-source work like DeepSeek-R1 demonstrates that reinforcement learning with verifiable reward is one key direction in reproducing o1. While the R1-style model has demonstrated success in language models, its application in multi-modal domains remains under-explored. This work introduces Visual Reinforcement Fine-Tuning (Visual-RFT), which further extends the application areas of RFT on visual tasks. Specifically, Visual-RFT first uses Large Vision-Language Models (LVLMs) to generate multiple responses containing reasoning tokens and final answers for each input, and then uses our proposed visual perception verifiable reward functions to update the model via the policy optimization algorithm such as Group Relative Policy Optimization (GRPO). We design different verifiable reward functions for different perception tasks, such as the Intersection over Union (IoU) reward for object detection. Experimental results on fine-grained image classification, few-shot object detection, reasoning grounding, as well as open-vocabulary object detection benchmarks show the competitive performance and advanced generalization ability of Visual-RFT compared with Supervised Fine-tuning (SFT). For example, Visual-RFT improves accuracy by $24.3\%$ over the baseline in one-shot fine-grained image classification with around 100 samples. In few-shot object detection, Visual-RFT also exceeds the baseline by $21.9$ on COCO's two-shot setting and $15.4$ on LVIS. Our Visual-RFT represents a paradigm shift in fine-tuning LVLMs, offering a data-efficient, reward-driven approach that enhances reasoning and adaptability for domain-specific tasks.
In 3D object mapping, category-level priors enable efficient object reconstruction and canonical pose estimation, requiring only a single prior per semantic category (e.g., chair, book, laptop). Recently, DeepSDF has predominantly been used as a category-level shape prior, but it struggles to reconstruct sharp geometry and is computationally expensive. In contrast, NeRFs capture fine details but have yet to be effectively integrated with category-level priors in a real-time multi-object mapping framework. To bridge this gap, we introduce PRENOM, a Prior-based Efficient Neural Object Mapper that integrates category-level priors with object-level NeRFs to enhance reconstruction efficiency while enabling canonical object pose estimation. PRENOM gets to know objects on a first-name basis by meta-learning on synthetic reconstruction tasks generated from open-source shape datasets. To account for object category variations, it employs a multi-objective genetic algorithm to optimize the NeRF architecture for each category, balancing reconstruction quality and training time. Additionally, prior-based probabilistic ray sampling directs sampling toward expected object regions, accelerating convergence and improving reconstruction quality under constrained resources. Experimental results on a low-end GPU highlight the ability of PRENOM to achieve high-quality reconstructions while maintaining computational feasibility. Specifically, comparisons with prior-free NeRF-based approaches on a synthetic dataset show a 21% lower Chamfer distance, demonstrating better reconstruction quality. Furthermore, evaluations against other approaches using shape priors on a noisy real-world dataset indicate a 13% improvement averaged across all reconstruction metrics, and comparable pose and size estimation accuracy, while being trained for 5x less time.
Recently, trimap-free methods have drawn increasing attention in human video matting due to their promising performance. Nevertheless, these methods still suffer from the lack of deterministic foreground-background cues, which impairs their ability to consistently identify and locate foreground targets over time and mine fine-grained details. In this paper, we present a trimap-free Object-Aware Video Matting (OAVM) framework, which can perceive different objects, enabling joint recognition of foreground objects and refinement of edge details. Specifically, we propose an Object-Guided Correction and Refinement (OGCR) module, which employs cross-frame guidance to aggregate object-level instance information into pixel-level detail features, thereby promoting their synergy. Furthermore, we design a Sequential Foreground Merging augmentation strategy to diversify sequential scenarios and enhance capacity of the network for object discrimination. Extensive experiments on recent widely used synthetic and real-world benchmarks demonstrate the state-of-the-art performance of our OAVM with only an initial coarse mask. The code and model will be available.
Using Quadrics as the object representation has the benefits of both generality and closed-form projection derivation between image and world spaces. Although numerous constraints have been proposed for dual quadric reconstruction, we found that many of them are imprecise and provide minimal improvements to localization.After scrutinizing the existing constraints, we introduce a concise yet more precise convex hull-based algebraic constraint for object landmarks, which is applied to object reconstruction, frontend pose estimation, and backend bundle adjustment.This constraint is designed to fully leverage precise semantic segmentation, effectively mitigating mismatches between complex-shaped object contours and dual quadrics.Experiments on public datasets demonstrate that our approach is applicable to both monocular and RGB-D SLAM and achieves improved object mapping and localization than existing quadric SLAM methods. The implementation of our method is available at https://github.com/tiev-tongji/convexhull-based-algebraic-constraint.
Room reidentification (ReID) is a challenging yet essential task with numerous applications in fields such as augmented reality (AR) and homecare robotics. Existing visual place recognition (VPR) methods, which typically rely on global descriptors or aggregate local features, often struggle in cluttered indoor environments densely populated with man-made objects. These methods tend to overlook the crucial role of object-oriented information. To address this, we propose AirRoom, an object-aware pipeline that integrates multi-level object-oriented information-from global context to object patches, object segmentation, and keypoints-utilizing a coarse-to-fine retrieval approach. Extensive experiments on four newly constructed datasets-MPReID, HMReID, GibsonReID, and ReplicaReID-demonstrate that AirRoom outperforms state-of-the-art (SOTA) models across nearly all evaluation metrics, with improvements ranging from 6% to 80%. Moreover, AirRoom exhibits significant flexibility, allowing various modules within the pipeline to be substituted with different alternatives without compromising overall performance. It also shows robust and consistent performance under diverse viewpoint variations.
Generative methods for image and video editing use generative models as priors to perform edits despite incomplete information, such as changing the composition of 3D objects shown in a single image. Recent methods have shown promising composition editing results in the image setting, but in the video setting, editing methods have focused on editing object's appearance and motion, or camera motion, and as a result, methods to edit object composition in videos are still missing. We propose \name as a method for editing 3D object compositions in videos of static scenes with camera motion. Our approach allows editing the 3D position of a 3D object across all frames of a video in a temporally consistent manner. This is achieved by lifting intermediate features of a generative model to a 3D reconstruction that is shared between all frames, editing the reconstruction, and projecting the features on the edited reconstruction back to each frame. To the best of our knowledge, this is the first generative approach to edit object compositions in videos. Our approach is simple and training-free, while outperforming state-of-the-art image editing baselines.
In egocentric video understanding, the motion of hands and objects as well as their interactions play a significant role by nature. However, existing egocentric video representation learning methods mainly focus on aligning video representation with high-level narrations, overlooking the intricate dynamics between hands and objects. In this work, we aim to integrate the modeling of fine-grained hand-object dynamics into the video representation learning process. Since no suitable data is available, we introduce HOD, a novel pipeline employing a hand-object detector and a large language model to generate high-quality narrations with detailed descriptions of hand-object dynamics. To learn these fine-grained dynamics, we propose EgoVideo, a model with a new lightweight motion adapter to capture fine-grained hand-object motion information. Through our co-training strategy, EgoVideo effectively and efficiently leverages the fine-grained hand-object dynamics in the HOD data. Extensive experiments demonstrate that our method achieves state-of-the-art performance across multiple egocentric downstream tasks, including improvements of 6.3% in EK-100 multi-instance retrieval, 5.7% in EK-100 classification, and 16.3% in EGTEA classification in zero-shot settings. Furthermore, our model exhibits robust generalization capabilities in hand-object interaction and robot manipulation tasks. Code and data are available at https://github.com/OpenRobotLab/EgoHOD/.
It remains a significant challenge to compress images at ultra-low bitrate while achieving both semantic consistency and high perceptual quality. We propose a novel image compression framework, Semantically Disentangled Image Compression (SEDIC) in this paper. Our proposed SEDIC leverages large multimodal models (LMMs) to disentangle the image into several essential semantic information, including an extremely compressed reference image, overall and object-level text descriptions, and the semantic masks. A multi-stage semantic decoder is designed to progressively restore the transmitted reference image object-by-object, ultimately producing high-quality and perceptually consistent reconstructions. In each decoding stage, a pre-trained controllable diffusion model is utilized to restore the object details on the reference image conditioned by the text descriptions and semantic masks. Experimental results demonstrate that SEDIC significantly outperforms state-of-the-art approaches, achieving superior perceptual quality and semantic consistency at ultra-low bitrates ($\le$ 0.05 bpp). Our code is available at https://github.com/yang-xidian/SEDIC.
This paper explores a cross-modality synthesis task that infers 3D human-object interactions (HOIs) from a given text-based instruction. Existing text-to-HOI synthesis methods mainly deploy a direct mapping from texts to object-specific 3D body motions, which may encounter a performance bottleneck since the huge cross-modality gap. In this paper, we observe that those HOI samples with the same interaction intention toward different targets, e.g., "lift a chair" and "lift a cup", always encapsulate similar action-specific body motion patterns while characterizing different object-specific interaction styles. Thus, learning effective action-specific motion priors and object-specific interaction priors is crucial for a text-to-HOI model and dominates its performances on text-HOI semantic consistency and body-object interaction realism. In light of this, we propose a novel body pose generation strategy for the text-to-HOI task: infer object-agnostic canonical body action first and then enrich object-specific interaction styles. Specifically, the first canonical body action inference stage focuses on learning intra-class shareable body motion priors and mapping given text-based semantics to action-specific canonical 3D body motions. Then, in the object-specific interaction inference stage, we focus on object affordance learning and enrich object-specific interaction styles on an inferred action-specific body motion basis. Extensive experiments verify that our proposed text-to-HOI synthesis system significantly outperforms other SOTA methods on three large-scale datasets with better semantic consistency and interaction realism performances.
Creating robots that can assist in farms and gardens can help reduce the mental and physical workload experienced by farm workers. We tackle the problem of object search in a farm environment, providing a method that allows a robot to semantically reason about the location of an unseen target object among a set of previously seen objects in the environment using a Large Language Model (LLM). We leverage object-to-object semantic relationships to plan a path through the environment that will allow us to accurately and efficiently locate our target object while also reducing the overall distance traveled, without needing high-level room or area-level semantic relationships. During our evaluations, we found that our method outperformed a current state-of-the-art baseline and our ablations. Our offline testing yielded an average path efficiency of 84%, reflecting how closely the predicted path aligns with the ideal path. Upon deploying our system on the Boston Dynamics Spot robot in a real-world farm environment, we found that our system had a success rate of 80%, with a success weighted by path length of 0.67, which demonstrates a reasonable trade-off between task success and path efficiency under real-world conditions. The project website can be viewed at https://adi-balaji.github.io/losae/
The free energy principle (FEP), along with the associated constructs of Markov blankets and ontological potentials, have recently been presented as the core components of a generalized modeling method capable of mathematically describing arbitrary objects that persist in random dynamical systems; that is, a mathematical theory of ``every'' ``thing''. Here, we leverage the FEP to develop a mathematical physics approach to the identification of objects, object types, and the macroscopic, object-type-specific rules that govern their behavior. We take a generative modeling approach and use variational Bayesian expectation maximization to develop a dynamic Markov blanket detection algorithm that is capable of identifying and classifying macroscopic objects, given partial observation of microscopic dynamics. This unsupervised algorithm uses Bayesian attention to explicitly label observable microscopic elements according to their current role in a given system, as either the internal or boundary elements of a given macroscopic object; and it identifies macroscopic physical laws that govern how the object interacts with its environment. Because these labels are dynamic or evolve over time, the algorithm is capable of identifying complex objects that travel through fixed media or exchange matter with their environment. This approach leads directly to a flexible class of structured, unsupervised algorithms that sensibly partition complex many-particle or many-component systems into collections of interacting macroscopic subsystems, namely, ``objects'' or ``things''. We derive a few examples of this kind of macroscopic physics discovery algorithm and demonstrate its utility with simple numerical experiments, in which the algorithm correctly labels the components of Newton's cradle, a burning fuse, the Lorenz attractor, and a simulated cell.
Neural synchrony is hypothesized to play a crucial role in how the brain organizes visual scenes into structured representations, enabling the robust encoding of multiple objects within a scene. However, current deep learning models often struggle with object binding, limiting their ability to represent multiple objects effectively. Inspired by neuroscience, we investigate whether synchrony-based mechanisms can enhance object encoding in artificial models trained for visual categorization. Specifically, we combine complex-valued representations with Kuramoto dynamics to promote phase alignment, facilitating the grouping of features belonging to the same object. We evaluate two architectures employing synchrony: a feedforward model and a recurrent model with feedback connections to refine phase synchronization using top-down information. Both models outperform their real-valued counterparts and complex-valued models without Kuramoto synchronization on tasks involving multi-object images, such as overlapping handwritten digits, noisy inputs, and out-of-distribution transformations. Our findings highlight the potential of synchrony-driven mechanisms to enhance deep learning models, improving their performance, robustness, and generalization in complex visual categorization tasks.
Hand-object interaction(HOI) is the fundamental link between human and environment, yet its dexterous and complex pose significantly challenges for gesture control. Despite significant advances in AI and robotics, enabling machines to understand and simulate hand-object interactions, capturing the semantics of functional grasping tasks remains a considerable challenge. While previous work can generate stable and correct 3D grasps, they are still far from achieving functional grasps due to unconsidered grasp semantics. To address this challenge, we propose an innovative two-stage framework, Functional Grasp Synthesis Net (FGS-Net), for generating 3D HOI driven by functional text. This framework consists of a text-guided 3D model generator, Functional Grasp Generator (FGG), and a pose optimization strategy, Functional Grasp Refiner (FGR). FGG generates 3D models of hands and objects based on text input, while FGR fine-tunes the poses using Object Pose Approximator and energy functions to ensure the relative position between the hand and object aligns with human intent and remains physically plausible. Extensive experiments demonstrate that our approach achieves precise and high-quality HOI generation without requiring additional 3D annotation data.
As smart homes become more prevalent in daily life, the ability to understand dynamic environments is essential which is increasingly dependent on AI systems. This study focuses on developing an intelligent algorithm which can navigate a robot through a kitchen, recognizing objects, and tracking their relocation. The kitchen was chosen as the testing ground due to its dynamic nature as objects are frequently moved, rearranged and replaced. Various techniques, such as SLAM feature-based tracking and deep learning-based object detection (e.g., Faster R-CNN), are commonly used for object tracking. Additionally, methods such as optical flow analysis and 3D reconstruction have also been used to track the relocation of objects. These approaches often face challenges when it comes to problems such as lighting variations and partial occlusions, where parts of the object are hidden in some frames but visible in others. The proposed method in this study leverages the YOLOv5 architecture, initialized with pre-trained weights and subsequently fine-tuned on a custom dataset. A novel method was developed, introducing a frame-scoring algorithm which calculates a score for each object based on its location and features within all frames. This scoring approach helps to identify changes by determining the best-associated frame for each object and comparing the results in each scene, overcoming limitations seen in other methods while maintaining simplicity in design. The experimental results demonstrate an accuracy of 97.72%, a precision of 95.83% and a recall of 96.84% for this algorithm, which highlights the efficacy of the model in detecting spatial changes.
Large vision-language models show tremendous potential in understanding visual information through human languages. However, they are prone to suffer from object hallucination, i.e., the generated image descriptions contain objects that do not exist in the image. In this paper, we reveal that object hallucination can be attributed to overconfidence in irrelevant visual features when soft visual tokens map to the LLM's word embedding space. Specifically, by figuring out the semantic similarity between visual tokens and LLM's word embedding, we observe that the smoothness of similarity distribution strongly correlates with the emergence of object hallucinations. To mitigate hallucinations, we propose using the Variational Information Bottleneck (VIB) to alleviate overconfidence by introducing stochastic noise, facilitating the constraining of irrelevant information. Furthermore, we propose an entropy-based noise-controlling strategy to enable the injected noise to be adaptively constrained regarding the smoothness of the similarity distribution. We adapt the proposed AdaVIB across distinct model architectures. Experimental results demonstrate that the proposed AdaVIB mitigates object hallucinations by effectively alleviating the overconfidence in irrelevant visual features, with consistent improvements on two object hallucination benchmarks.
Decomposing visual scenes into objects, as humans do, facilitates modeling object relations and dynamics. Object-Centric Learning (OCL) achieves this by aggregating image or video feature maps into object-level feature vectors, known as \textit{slots}. OCL's self-supervision via reconstructing the input from slots struggles with complex textures, thus many methods employ Vision Foundation Models (VFMs) to extract feature maps with better objectness. However, using VFMs merely as feature extractors does not fully unlock their potential. We propose Vector-Quantized VFMs for OCL (VQ-VFM-OCL, or VVO), where VFM features are extracted to facilitate object-level information aggregation and further quantized to strengthen supervision in reconstruction. Our VVO unifies OCL representatives into a concise architecture. Experiments demonstrate that VVO not only outperforms mainstream methods on object discovery tasks but also benefits downstream tasks like visual prediction and reasoning. The source code is available in the supplement.
Multi-view object tracking (MVOT) offers promising solutions to challenges such as occlusion and target loss, which are common in traditional single-view tracking. However, progress has been limited by the lack of comprehensive multi-view datasets and effective cross-view integration methods. To overcome these limitations, we compiled a Multi-View object Tracking (MVTrack) dataset of 234K high-quality annotated frames featuring 27 distinct objects across various scenes. In conjunction with this dataset, we introduce a novel MVOT method, Multi-View Integration Tracker (MITracker), to efficiently integrate multi-view object features and provide stable tracking outcomes. MITracker can track any object in video frames of arbitrary length from arbitrary viewpoints. The key advancements of our method over traditional single-view approaches come from two aspects: (1) MITracker transforms 2D image features into a 3D feature volume and compresses it into a bird's eye view (BEV) plane, facilitating inter-view information fusion; (2) we propose an attention mechanism that leverages geometric information from fused 3D feature volume to refine the tracking results at each view. MITracker outperforms existing methods on the MVTrack and GMTD datasets, achieving state-of-the-art performance. The code and the new dataset will be available at https://mii-laboratory.github.io/MITracker/.
Trajectory-based motion control has emerged as an intuitive and efficient approach for controllable video generation. However, the existing trajectory-based approaches are usually limited to only generating the motion trajectory of the controlled object and ignoring the dynamic interactions between the controlled object and its surroundings. To address this limitation, we propose a Chain-of-Thought-based motion controller for controllable video generation, named C-Drag. Instead of directly generating the motion of some objects, our C-Drag first performs object perception and then reasons the dynamic interactions between different objects according to the given motion control of the objects. Specifically, our method includes an object perception module and a Chain-of-Thought-based motion reasoning module. The object perception module employs visual language models to capture the position and category information of various objects within the image. The Chain-of-Thought-based motion reasoning module takes this information as input and conducts a stage-wise reasoning process to generate motion trajectories for each of the affected objects, which are subsequently fed to the diffusion model for video synthesis. Furthermore, we introduce a new video object interaction (VOI) dataset to evaluate the generation quality of motion controlled video generation methods. Our VOI dataset contains three typical types of interactions and provides the motion trajectories of objects that can be used for accurate performance evaluation. Experimental results show that C-Drag achieves promising performance across multiple metrics, excelling in object motion control. Our benchmark, codes, and models will be available at https://github.com/WesLee88524/C-Drag-Official-Repo.
Contrastive Language-Image Pre-training (CLIP) models excel in zero-shot classification, yet face challenges in complex multi-object scenarios. This study offers a comprehensive analysis of CLIP's limitations in these contexts using a specialized dataset, ComCO, designed to evaluate CLIP's encoders in diverse multi-object scenarios. Our findings reveal significant biases: the text encoder prioritizes first-mentioned objects, and the image encoder favors larger objects. Through retrieval and classification tasks, we quantify these biases across multiple CLIP variants and trace their origins to CLIP's training process, supported by analyses of the LAION dataset and training progression. Our image-text matching experiments show substantial performance drops when object size or token order changes, underscoring CLIP's instability with rephrased but semantically similar captions. Extending this to longer captions and text-to-image models like Stable Diffusion, we demonstrate how prompt order influences object prominence in generated images. For more details and access to our dataset and analysis code, visit our project repository: https://clip-analysis.github.io.