3D semantic field learning is crucial for applications like autonomous navigation, AR/VR, and robotics, where accurate comprehension of 3D scenes from limited viewpoints is essential. Existing methods struggle under sparse view conditions, relying on inefficient per-scene multi-view optimizations, which are impractical for many real-world tasks. To address this, we propose SLGaussian, a feed-forward method for constructing 3D semantic fields from sparse viewpoints, allowing direct inference of 3DGS-based scenes. By ensuring consistent SAM segmentations through video tracking and using low-dimensional indexing for high-dimensional CLIP features, SLGaussian efficiently embeds language information in 3D space, offering a robust solution for accurate 3D scene understanding under sparse view conditions. In experiments on two-view sparse 3D object querying and segmentation in the LERF and 3D-OVS datasets, SLGaussian outperforms existing methods in chosen IoU, Localization Accuracy, and mIoU. Moreover, our model achieves scene inference in under 30 seconds and open-vocabulary querying in just 0.011 seconds per query.
论文检索
输入标题、作者或关键词，从 2,942 篇学术成果中精准定位
Creating digitalized hand-object interaction scenes plays a crucial role in recent advancements, enabling viewers to understand how human dexterity influences and shapes the world. In this paper, we present HandCraft, a framework designed to capture and render hand-object interactions with exceptional precision and realism. Our Gaussian models are built on the development of digital representations of hands, objects, and scenes, derived from data captured using multi-modal sensing systems. By combining motion capture with IMU-based data gloves equipped with tactile sensors, HandCraft ensures precise hand pose tracking and reliable contact fidelity. HandCraft includes a novel method that uses hand motions to solve the object occlusions, effectively reconstructing missing interaction details. For enhanced physical feasibility, HandCraft incorporates optimization techniques to resolve object penetration issues and enforce temporal consistency. Using these techniques, we introduce a high-quality dataset of hand-object interaction sequences, featuring complex and prolonged daily activities. This dataset demonstrates HandCraft's ability to capture and reproduce subtle, dynamic interactions in rich detail. HandCraft holds promises in creating realistic virtual environments and advancing world modeling in both graphics and robotics research.
In robotics and autonomous driving, accurate depth estimation is vital yet challenging under dynamic scenes and extreme lighting. Conventional frame-based cameras offer rich context but suffer from motion blur and limited dynamic range, while event cameras provide high temporal resolution and dynamic range but lack global scene structure. Therefore, recent studies explore frame-event fusion depth estimation methods to leverage these two complementary modalities to achieve robust performance. However, due to the mismatch in temporal and spatial resolution, there is an inherent contradiction between high spatial resolution frames captured at sparse temporal intervals and event streams characterized by spatial sparsity but high temporal resolution, rendering cross-modal feature fusion ineffective. Moreover, the limited availability of frame-event depth datasets further undermines the model's generalization capability across different scenes. To address the above challenges, we propose HAFUNet, a Hierarchical Attention Fusion Network for depth estimation via frame-event fusion. Our method contains: (1) a pre-trained Dual-Stream Encoder (DSEer) to extract complementary features from frame and event inputs; (2) a Cross-modal Feature Interaction Module (CFIM) that aligns and fuses spatial-channel features across modalities; and (3) a Hierarchical Attention Decoder (HADer) that progressively refines depth predictions via attention-guided convolution. Experiments on synthetic and real-world datasets show that HAFUNet surpasses existing methods in depth accuracy and robustness. These results demonstrate the strength of our fusion strategy in diverse environments. Code is available at https://github.com/SiYZhangwh/HAFUNet.
Open-vocabulary 3D visual grounding aims to localize target objects based on free-form language queries, which is crucial for embodied AI applications such as autonomous navigation, robotics, and augmented reality. Learning 3D language fields through neural representations enables accurate understanding of 3D scenes from limited viewpoints and facilitates the localization of target objects in complex environments. However, existing language field methods struggle to accurately localize instances using spatial relations in language queries, such as ''the book on the chair.'' This limitation mainly arises from inadequate reasoning about spatial relations in both language queries and 3D scenes. In this work, we propose SpatialReasoner, a novel neural representation-based framework with large language model (LLM)-driven spatial reasoning that constructs a visual properties-enhanced hierarchical feature field for open-vocabulary 3D visual grounding. To enable spatial reasoning in language queries, SpatialReasoner fine-tunes an LLM to capture spatial relations and explicitly infer instructions for the target, anchor, and spatial relation. To enable spatial reasoning in 3D scenes, SpatialReasoner incorporates visual properties (opacity and color) to construct a hierarchical feature field. This field represents language and instance features using distilled CLIP features and masks extracted via the Segment Anything Model (SAM). The field is then queried using the inferred instructions in a hierarchical manner to localize the target 3D instance based on the spatial relation in the language query. Notably, SpatialReasoner is not limited to a specific 3D neural representation; it serves as a framework adaptable to various representations, such as Neural Radiance Fields (NeRF) or 3D Gaussian Splatting (3DGS). Extensive experiments show that our framework can be seamlessly integrated into different neural representations, outperforming baseline models in 3D visual grounding while empowering their spatial reasoning capability. Project Homepage:ZhenyangLiu.github.io/SpatialReasoner.
Human behavior has the nature of mutual dependencies, which requires human-robot interactive systems to predict surrounding agents' trajectories by modeling complex social interactions, avoiding collisions and executing safe path planning. While there exist many trajectory prediction methods, most of them do not incorporate the own motion of the ego agent and only model interactions based on static information. We are inspired by the humans' theory of mind during trajectory selection and propose a Cross time domain intention-interactive method for conditional Trajectory prediction(CiT). Our proposed CiT conducts joint analysis of behavior intentions over time, and achieves information complementarity and integration across different time domains. The intention in its own time domain can be corrected by the social interaction information from the other time domain to obtain a more precise intention representation. In addition, CiT is designed to closely integrate with robotic motion planning and control modules, capable of generating a set of optional trajectory prediction results for all surrounding agents based on potential motions of the ego agent. Extensive experiments demonstrate that the proposed CiT significantly outperforms the existing methods, achieving state-of-the-art performance in the benchmarks.
Around 1990, the Berlin Wall came down, the WWW was born at CERN, mobile phones became popular, self-driving cars appeared in traffic, and modern AI based on very deep artificial neural networks emerged, including the principles behind the G, P, and T in ChatGPT. I place these events in the history of the universe since the Big Bang, and discuss what's next: not just AI behind the screen in the virtual world, but real AI for real robots in the real world, connected through a WWW of machines. Intelligent (but not necessarily super-intelligent) robots that can learn to operate the tools and machines operated by humans can also build (and repair when needed) more of their own kind. This will culminate in life-like, self-replicating and self-improving machine civilisations, which represent the ultimate form of upscaling, and will shape the long-term future of the entire cosmos. The wonderful short-term side effect is that our AI will continue to make people's lives longer, healthier and easier.
Unmanned Aerial Vehicles (UAVs) and Automated Guided Vehicles (AGVs) increasingly collaborate in logistics, surveillance, inspection tasks and etc. However, existing simulators often focus on a single domain, limiting cross-domain study. This paper presents the SkyRover, a modular simulator for UAV-AGV multi-agent pathfinding (MAPF). SkyRover supports realistic agent dynamics, configurable 3D environments, and convenient APIs for external solvers and learning methods. By unifying ground and aerial operations, it facilitates cross-domain algorithm design, testing, and benchmarking. Experiments highlight SkyRover’s capacity for efficient pathfinding and high-fidelity simulations in UAV-AGV coordination. We believe the SkyRover fills a key gap in MAPF research. Project is available at https://sites.google.com/view/mapf3d/home.
Reward Adaptation via Q-Manipulation: Provably Beneficial Reward Function Transfer in Reinforcement Learning
PDF ↗Reinforcement Learning has made great strides in game playing and robotics but faces challenges with sample complexity and generalization. Transfer learning, which allows agents to reuse knowledge from prior tasks, offers a promising solution. My current research focuses on Reward Adaptation, where agents adjust to new reward functions while leveraging knowledge from tasks with different reward functions. I propose Q-Manipulation (Q-M), a method that adapts Q-functions to new rewards by computing and iteratively tightening bounds, akin to value iteration. This allows for action pruning before learning begins, enhancing sample efficiency without compromising policy optimality. Through empirical comparisons I demonstrate its effectiveness, generalizability, and practicality. Future work will handle changes in transition dynamics and continuous MDPs.
3D spatial understanding is essential in real-world applications such as robotics, autonomous vehicles, virtual reality, and medical imaging. Recently, Large Language Models (LLMs), having demonstrated remarkable success across various domains, have been leveraged to enhance 3D understanding tasks, showing potential to surpass traditional computer vision methods. In this survey, we present a comprehensive review of methods integrating LLMs with 3D spatial understanding. We propose a taxonomy that categorizes existing methods into three branches: image-based methods deriving 3D understanding from 2D visual data, point cloud-based methods working directly with 3D representations, and hybrid modality-based methods combining multiple data streams. We systematically review representative methods along these categories, covering data representations, architectural modifications, and training strategies that bridge textual and 3D modalities. Finally, we discuss current limitations, including dataset scarcity and computational challenges, while highlighting promising research directions in spatial perception, multi-modal fusion, and real-world applications.
Embodied multi-agent systems (EMAS) have attracted growing attention for their potential to address complex, real-world challenges in areas such as logistics and robotics. Recent advances in foundation models pave the way for generative agents capable of richer communication and adaptive problem-solving. This survey provides a systematic examination of how EMAS can benefit from these generative capabilities. We propose a taxonomy that categorizes EMAS by system architectures and embodiment modalities, emphasizing how collaboration spans both physical and virtual contexts. Central building blocks, perception, planning, communication, and feedback, are then analyzed to illustrate how generative techniques bolster system robustness and flexibility. Through concrete examples, we demonstrate the transformative effects of integrating foundation models into embodied, multi-agent frameworks. Finally, we discuss challenges and future directions, underlining the significant promise of EMAS to reshape the landscape of AI-driven collaboration.
A Comprehensive Survey on Physical Risk Control in the Era of Foundation Model-enabled Robotics
PDF ↗Recent Foundation Model-enabled robotics (FMRs) display greatly improved general-purpose skills, enabling more adaptable automation than conventional robotics. Their ability to handle diverse tasks thus creates new opportunities to replace human labor. However, unlike general foundation models, FMRs interact with the physical world, where their actions directly affect the safety of humans and surrounding objects, requiring careful deployment and control. Based on this proposition, our survey comprehensively summarizes robot control approaches to mitigate physical risks by covering all the lifespan of FMRs ranging from pre-deployment to post-accident stage. Specifically, we broadly divide the timeline into the following three phases: (1) pre-deployment phase, (2) pre-incident phase, and (3) post-incident phase. Throughout this survey, we find that there is much room to study (i) pre-incident risk mitigation strategies, (ii) research that assumes physical interaction with humans, and (iii) essential issues of foundation models themselves. We hope that this survey will be a milestone in providing a high-resolution analysis of the physical risks of FMRs and their control, contributing to the realization of a good human-robot relationship.
Accurate trajectory prediction has long been a major challenge for autonomous driving (AD). Traditional data-driven models predominantly rely on statistical correlations, often overlooking the causal relationships that govern traffic behavior. In this paper, we introduce a novel trajectory prediction framework that leverages causal inference to enhance predictive robustness, generalization, and accuracy. By decomposing the environment into spatial and temporal components, our approach identifies and mitigates spurious correlations, uncovering genuine causal relationships. We also employ a progressive fusion strategy to integrate multimodal information, simulating human-like reasoning processes and enabling real-time inference. Evaluations on five real-world datasets—ApolloScape, nuScenes, NGSIM, HighD, and MoCAD—demonstrate our model's superiority over existing state-of-the-art (SOTA) methods, with improvements in key metrics such as RMSE and FDE. Our findings highlight the potential of causal reasoning to transform trajectory prediction, paving the way for robust AD systems.
AI and related technologies are reshaping jobs and tasks, either by automating or augmenting human skills in the workplace. Many researchers have been working on estimating if and to what extent jobs and tasks are exposed to the risk of being automatized by AI-related technologies. Our work tackles this issue through a data-driven approach by: (i) developing a reproducible framework that uses cutting-edge open-source large language models to assess the current capabilities of AI and robotics in performing job-related tasks; (ii) formalizing and computing a measure of AI exposure by occupation, the Task Exposure to AI (TEAI) index, and a measure of Task Replacement by AI (TRAI) index, both validated through a human user evaluation and compared with the state-of-the-art. Our results show that the TEAI index is positively correlated with cognitive, problem-solving, and management skills, while it is negatively correlated with social skills. Results also suggest that about one-third of U.S. employment is highly exposed to AI, primarily in high-skill jobs requiring a graduate or postgraduate level of education. We also find that AI exposure is positively associated with employment and wage growth from 2003 to 2023, suggesting that AI has had an overall positive effect on productivity. Considering specifically the TRAI index, we find that even in high-skill occupations, AI exhibits high variability in task substitution, suggesting that AI and humans complement each other within the same occupation, while the allocation of tasks within occupations is likely to change. All results, models, and code are freely available online to allow the community to reproduce our results, compare outcomes, and use our work as a benchmark to monitor AI’s progress over time.
Black-Box Optimization (BBO) has found successful applications in many fields of science and engineering. Recently, there has been a growing interest in meta-learning particular components of BBO algorithms to speed up optimization and get rid of tedious hand-crafted heuristics. As an extension, learning the entire algorithm from data requires the least labor from experts and can provide the most flexibility. In this paper, we propose RIBBO, a method to reinforce-learn a BBO algorithm from offline data in an end-to-end fashion. RIBBO employs expressive sequence models to learn the optimization histories produced by multiple behavior algorithms and tasks, leveraging the in-context learning ability of large models to extract task information and make decisions accordingly. Central to our method is to augment the optimization histories with regret-to-go tokens, which are designed to represent the performance of an algorithm based on cumulative regret over the future part of the histories. The integration of regret-to-go tokens enables RIBBO to automatically generate sequences of query points that are positively correlated to the user-desired regret, verified by its universally good empirical performance on diverse problems, including BBO benchmark, hyper-parameter optimization, and robot control problems.
Theoretical Analysis of Evolutionary Algorithms with Quality Diversity for a Classical Path Planning Problem
PDF ↗Quality diversity (QD) algorithms, an extension of evolutionary algorithms, excel at generating diverse sets of high-quality solutions for complex problems in robotics, games, and combinatorial optimisation. Despite their success, the underlying mechanisms remain poorly understood due to a lack of a theoretical foundation. We address this gap by analysing QD algorithms on the all-pairs-shortest-paths (APSP) problem, a classical planning task that naturally seeks multiple solutions. Using Map-Elites, a prominent QD approach, we leverage its ability to evolve solutions across distinct regions of a behavioural space, which for APSP corresponds to all pairs of nodes in the graph. Our analysis rigorously demonstrates that evolutionary algorithms using Map-Elites efficiently compute shortest paths for all node pairs in parallel by exploiting synergies in the behavioural space. By appending edges to an existing shortest path, mutation can create optimal solutions in other regions of the behavioural space. Crossover is particularly effective, as it can combine optimal paths from two regions to produce an optimal path for a third region simply by concatenating two shortest paths. Finally, refining the parent selection to facilitate successful crossovers exhibits significant speed-ups compared to standard QD approaches.
Self-supervised monocular depth estimation has attracted significant attention due to its broad applications in autonomous driving and robotics. Although significant performance improvements have been achieved by learning the relative distance of objects with the introduction of Self Query Layer (SQL), it struggles with zero-shot generalization due to the lack of geometric features and the fixed number of query sizes. To address these problems, we propose a diffusion-augmented self-supervised depth estimation framework, named DiffSQL, to learn geometric priors for feature augmentation. Additionally, we introduce a dynamic self-query layer that implicitly computes the relative distances between objects by adjusting the query size according to the feature distribution. Experimental results on the KITTI dataset show that DiffSQL outperforms SQLdepth by 1.03% in terms of AbsRel and 2.79% in terms of SqRel. Furthermore, our experiments demonstrate that DiffSQL is superior in zero-shot generalization.
Code-BT: A Code-Driven Approach to Behavior Tree Generation for Robot Tasks Planning with Large Language Models
PDF ↗Behavior trees(BTs) provide a systematic and structured control architecture extensively employed in game AI and robotic behavior control, owing to their modularity, reactivity, and reusability. Nonetheless, manual BTs design requires significant expertise and becomes inefficient as task complexity increases. Recent automation technologies have avoided manual work, but often have high application barriers and face challenges in adapting to new tasks, making it difficult to easily configure them to specific requirements. Code-BT introduces a novel approach that utilizes large language models(LLMs) to automatically generate BTs, representing the task planning process as the process of coding and organizing sequences. By retrieving control flow information from the generated code, BTs can be efficiently constructed to address the complexity and diversity of task planning challenges. Rather than relying on manual design, Code-BT uses task instructions to guide the selection of relevant APIs, and then systematically assembles these APIs into modular code to align with the BTs structure. Finally, action sequences and control logic are extracted from the generated code to construct the BTs. Our approach not only ensures the automation of BTs generation but also guarantees the scalability and adaptability for long-term tasks. Experimental results demonstrate that Code-BT substantially improves LLM performance in BTs generation, achieving improvements ranging from16.67% to 29.17%.
Point cloud extraction (PCE) and ego velocity estimation (EVE) are key capabilities gaining attention in 3D radar perception. However, existing work typically treats these two tasks independently, which may neglect the interplay between radar's spatial and Doppler domain features, potentially introducing additional bias. In this paper, we observe an underlying correlation between 3D points and ego velocity, which offers reciprocal benefits for PCE and EVE. To fully unlock such inspiring potential, we take the first step to design a Spatial-Doppler Diffusion (SDDiff) model for simultaneously dense PCE and accurate EVE. To seamlessly tailor it to radar perception, SDDiff improves the conventional latent diffusion process in three major aspects. First, we introduce a representation that embodies both spatial occupancy and Doppler features. Second, we design a directional diffusion with radar priors to streamline the sampling. Third, we propose Iterative Doppler Refinement to enhance the model’s adaptability to density variations and ghosting effects. Extensive evaluations show that SDDiff significantly outperforms state-of-the-art baselines by achieving 59% higher in EVE accuracy, 4X greater in valid generation density while boosting PCE effectiveness and reliability. The code and dataset will be available on https://github.com/StellarEsti/SDDiff.
We introduce DisPIM, a framework that leverages pretrained image models (PIMs) for visuo-motor control. Applying PIMs to visuo-motor control faces a big difficulty due to the distribution shift between the distribution of visual environmental states and that of the pretraining datasets. Due to such a distribution shift, fine-tuning PIMs specifically for visuo-motor control may hurt the generalizability of PIMs, while adding additional tunable parameters for specific actions apparently lead to high computational costs. DisPIM addresses these challenges using a novel feature distillation approach, which obtains a compact model that not only inherit the generalization capability of PIMs but also acquire task-specific skills for visuo-motor control. This good for both sides is mainly achieved by means of a target Q-ensemble mechanism, which is inspired by double Q-learning. This Q-ensemble mechanism can adaptively adjust the distillation rate, so as to balance the objective of generalization and task-specific ability during training. With this balancing mechanism, DisPIM achieves both task-specific and generalizable control requiring a low computation cost. Across a series of algorithms, task domains, and evaluation metrics in both simulation and real robot, our DisPIM demonstrates significant improvements in generalization and overall performance with low computational overhead.
Inertial navigation enables self-contained localization using only Inertial Measurement Units (IMUs), making it widely applicable in various domains such as navigation, augmented reality, and robotics. However, existing methods suffer from drift accumulation due to the sensor noise and difficulty capturing long-range temporal dependencies, limiting their robustness and accuracy. To address these challenges, we propose DiffusionIMU, a novel diffusion-based framework for inertial navigation. DiffusionIMU enhances direct velocity regression from IMU data through an iterative generative denoising process, progressively refining motion state estimation. It integrates the noise-adaptive feature modulation for sensor variability handling, the feature alignment mechanism for representation consistency, and the diffusion-based temporal modeling to decrease accumulated drift. Experiments show that DiffusionIMU consistently outperforms existing methods, demonstrating superior generalization to unseen users while alleviating the impact of the sensor noise.