AI Agents

Visual reasoning -- the ability to interpret the visual world -- is crucial for embodied agents that operate within three-dimensional scenes. Progress in AI has led to vision and language models capable of answering questions from images. However, their performance declines when tasked with 3D spatial reasoning. To tackle the complexity of such reasoning problems, we introduce an agentic program synthesis approach where LLM agents collaboratively generate a Pythonic API with new functions to solve common subproblems. Our method overcomes limitations of prior approaches that rely on a static, human-defined API, allowing it to handle a wider range of queries. To assess AI capabilities for 3D understanding, we introduce a new benchmark of queries involving multiple steps of grounding and inference. We show that our method outperforms prior zero-shot models for visual reasoning in 3D and empirically validate the effectiveness of our agentic framework for 3D spatial reasoning tasks. Project website: https://glab-caltech.github.io/vadar/

Video Large Language Models (Video LLMs) have recently exhibited remarkable capabilities in general video understanding.However, they mainly focus on holistic comprehension and struggle with capturing fine-grained spatial and temporal details. Besides, the lack of high-quality object-level video instruction data and a comprehensive benchmark further hinders their advancements. To tackle these challenges, we introduce the VideoRefer Suite to empower Video LLM for finer-level spatial-temporal video understanding, i.e., enabling perception and reasoning on any objects throughout the video. Specially, we thoroughly develop VideoRefer Suite across three essential aspects: dataset, model, and benchmark. Firstly, we introduce a multi-agent data engine to meticulously curate a large-scale, high-quality object-level video instruction dataset, termed VideoRefer-700K. Next, we present the VideoRefer model, which equips a versatile spatial-temporal object encoder to capture precise regional and sequential representations. Finally, we meticulously create a VideoRefer-Bench to comprehensively assess the spatial-temporal understanding capability of a Video LLM, evaluating it across various aspects. Extensive experiments and analyses demonstrate that our VideoRefer model not only achieves promising performance on video referring benchmarks but also facilitates general video understanding capabilities. Our codes, models, and dataset will be made publicly available.

Multi-agent trajectory modeling has primarily focused on forecasting future states, often overlooking broader tasks like trajectory completion, which are crucial for real-world applications such as correcting tracking data. Existing methods also generally predict agents' states without offering any state-wise measure of uncertainty. Moreover, popular multi-modal sampling methods lack any error probability estimates for each generated scene under the same prior observations, making it difficult to rank the predictions during inference time. We introduce U2Diff, a unified diffusion model designed to handle trajectory completion while providing state-wise uncertainty estimates jointly. This uncertainty estimation is achieved by augmenting the simple denoising loss with the negative log-likelihood of the predicted noise and propagating latent space uncertainty to the real state space. Additionally, we incorporate a Rank Neural Network in post-processing to enable error probability estimation for each generated mode, demonstrating a strong correlation with the error relative to ground truth. Our method outperforms the state-of-the-art solutions in trajectory completion and forecasting across four challenging sports datasets (NBA, Basketball-U, Football-U, Soccer-U), highlighting the effectiveness of uncertainty and error probability estimation.

Cooperative perception presents significant potential for enhancing the sensing capabilities of individual vehicles, however, inter-agent latency remains a critical challenge. Latencies cause misalignments in both spatial and semantic features, complicating the fusion of real-time observations from the ego vehicle with delayed data from others. To address these issues, we propose TraF-Align, a novel framework that learns the flow path of features by predicting the feature-level trajectory of objects from past observations up to the ego vehicle's current time. By generating temporally ordered sampling points along these paths, TraF-Align directs attention from the current-time query to relevant historical features along each trajectory, supporting the reconstruction of current-time features and promoting semantic interaction across multiple frames. This approach corrects spatial misalignment and ensures semantic consistency across agents, effectively compensating for motion and achieving coherent feature fusion. Experiments on two real-world datasets, V2V4Real and DAIR-V2X-Seq, show that TraF-Align sets a new benchmark for asynchronous cooperative perception.

Large Language Models (LLMs) have demonstrated excellent capabilities in composing various modules together to create programs that can perform complex reasoning tasks on images. In this paper, we propose TANGO, an approach that extends the program composition via LLMs already observed for images, aiming to integrate those capabilities into embodied agents capable of observing and acting in the world. Specifically, by employing a simple PointGoal Navigation model combined with a memory-based exploration policy as a foundational primitive for guiding an agent through the world, we show how a single model can address diverse tasks without additional training. We task an LLM with composing the provided primitives to solve a specific task, using only a few in-context examples in the prompt. We evaluate our approach on three key Embodied AI tasks: Open-Set ObjectGoal Navigation, Multi-Modal Lifelong Navigation, and Open Embodied Question Answering, achieving state-of-the-art results without any specific fine-tuning in challenging zero-shot scenarios.

Graphical User Interface (GUI) agents demonstrate promising potential in assisting human-computer interaction, automating human user's navigation on digital devices. An ideal GUI agent is expected to achieve high accuracy, low latency, and compatibility for different GUI platforms. Recent vision-based approaches have shown promise by leveraging advanced Vision Language Models (VLMs). While they generally meet the requirements of compatibility and low latency, these vision-based GUI agents tend to have low accuracy due to their limitations in element grounding. To address this issue, we propose SpiritSight, a vision-based, end-to-end GUI agent that excels in GUI navigation tasks across various GUI platforms. First, we create a multi-level, large-scale, high-quality GUI dataset called GUI-Lasagne using scalable methods, empowering SpiritSight with robust GUI understanding and grounding capabilities. Second, we introduce the Universal Block Parsing (UBP) method to resolve the ambiguity problem inherited from the dynamic resolution strategy, further enhancing SpiritSight's ability to ground GUI objects. Through these efforts, SpiritSight agent outperforms other advanced methods on diverse GUI benchmarks, demonstrating its superior capability and compatibility in GUI navigation tasks. The models and code will be made available upon publication.

The analysis and prediction of agent trajectories are crucial for decision-making processes in intelligent systems, with precise short-term trajectory forecasting being highly significant across a range of applications. Agents and their social interactions have been quantified and modeled by researchers from various perspectives; however, substantial limitations exist in the current work due to the inherent high uncertainty of agent intentions and the complex higher-order influences among neighboring groups. SocialMOIF is proposed to tackle these challenges, concentrating on the higher-order intention interactions among neighboring groups while reinforcing the primary role of first-order intention interactions between neighbors and the target agent. This method develops a multi-order intention fusion model to achieve a more comprehensive understanding of both direct and indirect intention information. Within SocialMOIF, a trajectory distribution approximator is designed to guide the trajectories toward values that align more closely with the actual data, thereby enhancing model interpretability. Furthermore, a global trajectory optimizer is introduced to enable more accurate and efficient parallel predictions. By incorporating a novel loss function that accounts for distance and direction during training, experimental results demonstrate that the model outperforms previous state-of-the-art baselines across multiple metrics in both dynamic and static datasets.

We introduce Sketchtopia, a large-scale dataset and AI framework designed to explore goal-driven, multimodal communication through asynchronous interactions in a Pictionary-inspired setup. Sketchtopia captures natural human interactions, including freehand sketches, open-ended guesses, and iconic feedback gestures, showcasing the complex dynamics of cooperative communication under constraints. It features over 20K gameplay sessions from 916 players, capturing 263K sketches, 10K erases, 56K guesses and 19.4K iconic feedbacks. We introduce multimodal foundational agents with capabilities for generative sketching, guess generation and asynchronous communication. Our dataset also includes 800 human-agent sessions for benchmarking the agents. We introduce novel metrics to characterize collaborative success, responsiveness to feedback and inter-agent asynchronous communication. Sketchtopia pushes the boundaries of multimodal AI, establishing a new benchmark for studying asynchronous, goal-oriented interactions between humans and AI agents. The dataset can be found at https://sketchtopia25.github.io/

Modeling spatial-temporal interactions among neighboring agents is at the heart of multi-agent problems such as motion forecasting and crowd navigation. Despite notable progress, it remains unclear to which extent modern representations can capture the causal relationships behind agent interactions. In this work, we take an in-depth look at the causal awareness of these representations, from computational formalism to real-world practice. First, we revisit the notion of non-causal robustness studied in the recent CausalAgents benchmark. We show that existing representations are already partially resilient to perturbations of non-causal agents, and yet modeling indirect causal effects involving mediator agents remains challenging. To address this challenge, we introduce a metric learning approach that regularizes latent representations with causal annotations. Our controlled experiments show that this approach not only leads to higher degrees of causal awareness but also yields stronger out-of-distribution robustness. To further operationalize it in practice, we propose a sim-to-real causal transfer method via cross-domain multi-task learning. Experiments on trajectory prediction datasets show that our method can significantly boost generalization, even in the absence of real-world causal annotations, where we acquire higher prediction accuracy by only using 25% of real-world data. We hope our work provides a new perspective on the challenges and potential pathways toward causally-aware representations of multi-agent interactions. Our code is available in supplementary materials. Our code is available at https://github.com/vita-epfl/CausalSim2Real.

Progress in Embodied AI has made it possible for end-to-end-trained agents to navigate in photo-realistic environments with high-level reasoning and zero-shot or language-conditioned behavior, but evaluations and benchmarks are still dominated by simulation. In this work, we focus on the fine-grained behavior of fast-moving real robots and present a large-scale experimental study involving \numepisodes navigation episodes in a real environment with a physical robot, where we analyze the type of reasoning emerging from end-to-end training. In particular, we study the presence of realistic dynamics which the agent learned for open-loop forecasting, and their interplay with sensing. We analyze the way the agent uses latent memory to hold elements of the scene structure and information gathered during exploration. We probe the planning capabilities of the agent, and find in its memory evidence for somewhat precise plans over a limited horizon. Furthermore, we show in a post-hoc analysis that the value function learned by the agent relates to long-term planning. Put together, our experiments paint a new picture on how using tools from computer vision and sequential decision making have led to new capabilities in robotics and control. An interactive tool is available at https://visual-navigation-reasoning.github.io

We introduce a simple framework for predicting the behavior of an agent in multi-agent settings. In contrast to autoregressive (AR) tasks, such as language processing, our focus is on scenarios with multiple agents whose interactions are shaped by physical constraints and internal motivations. To this end, we propose Poly-Autoregressive (PAR) modeling, which forecasts an ego agent's future behavior by reasoning about the ego agent's state history and the past and current states of other interacting agents. At its core, PAR represents the behavior of all agents as a sequence of tokens, each representing an agent's state at a specific timestep. With minimal data pre-processing changes, we show that PAR can be applied to three different problems: human action forecasting in social situations, trajectory prediction for autonomous vehicles, and object pose forecasting during hand-object interaction. Using a small proof-of-concept transformer backbone, PAR outperforms AR across these three scenarios.

High-speed autofocus in extreme scenes remains a significant challenge. Traditional methods rely on repeated sampling around the focus position, resulting in "focus hunting". Event-driven methods have advanced focusing speed and improved performance in low-light conditions; however, current approaches still require at least one lengthy round of "focus hunting", involving the collection of a complete focus stack. We introduce the Event Laplacian Product (ELP) focus detection function, which combines event data with grayscale Laplacian information, redefining focus search as a detection task. This innovation enables the first one-step event-driven autofocus, cutting focusing time by up to two-thirds and reducing focusing error by 24 times on the DAVIS346 dataset and 22 times on the EVK4 dataset. Additionally, we present an autofocus pipeline tailored for event-only cameras, achieving accurate results across a range of challenging motion and lighting conditions. All datasets and code will be made publicly available.

The advances in vision-language models (VLMs) have led to a growing interest in autonomous driving to leverage their strong reasoning capabilities. However, extending these capabilities from 2D to full 3D understanding is crucial for real-world applications. To address this challenge, we propose OmniDrive, a holistic vision-language dataset that aligns agent models with 3D driving tasks through counterfactual reasoning. This approach enhances decision-making by evaluating potential scenarios and their outcomes, similar to human drivers considering alternative actions. Our counterfactual-based synthetic data annotation process generates large-scale, high-quality datasets, providing denser supervision signals that bridge planning trajectories and language-based reasoning. Futher, we explore two advanced OmniDrive-Agent frameworks, namely Omni-L and Omni-Q, to assess the importance of vision-language alignment versus 3D perception, revealing critical insights into designing effective LLM-agents. Significant improvements on the DriveLM Q&A benchmark and nuScenes open-loop planning demonstrate the effectiveness of our dataset and methods.

Designing effective neural architectures poses a significant challenge in deep learning. While Neural Architecture Search (NAS) automates the search for optimal architectures, existing methods are often constrained by predetermined search spaces and may miss critical neural architectures. In this paper, we introduce NADER (Neural Architecture Design via multi-agEnt collaboRation), a novel framework that formulates neural architecture design (NAD) as a LLM-based multi-agent collaboration problem. NADER employs a team of specialized agents to enhance a base architecture through iterative modification. Current LLM-based NAD methods typically operate independently, lacking the ability to learn from past experiences, which results in repeated mistakes and inefficient exploration. To address this issue, we propose the Reflector, which effectively learns from immediate feedback and long-term experiences. Additionally, unlike previous LLM-based methods that use code to represent neural architectures, we utilize a graph-based representation. This approach allows agents to focus on design aspects without being distracted by coding. We demonstrate the effectiveness of NADER in discovering high-performing architectures beyond predetermined search spaces through extensive experiments on benchmark tasks, showcasing its advantages over state-of-the-art methods.

We present Magma, a foundation model that serves multimodal AI agentic tasks in both the digital and physical worlds. Magma is a significant extension of vision-language (VL) models in that it not only retains the VL understanding ability (verbal intelligence) of the latter, but is also equipped with the ability to ground and act in the visual-spatial world (spatial-temporal intelligence). To endow agentic capabilities for tasks ranging from UI navigation to robot manipulation, Magma is trained on large amounts of heterogeneous datasets that span from images, videos to robotics data, where actionable visual objects (e.g. clickable buttons in GUI) in images are labeled by Set-of-Mark (SoM) for action grounding, and object movements (e.g. trace of human hands or robotic arms) in videos are labeled by Trace-of-Mark (ToM) for action planning. Extensive experiments show that SoM and ToM help bridge the gap between verbal and action abilities and significantly enhance spatio-temporal intelligence which is fundamental to agentic tasks, as shown in Fig.1. In particular, Magma creates new state-of-the-art results on UI navigation and robotic manipulation tasks, outperforming previous models that are specifically tailored to these tasks. Moreover, Magma preserves strong multimodal understanding ability and compares favorably to popular large multimodal models that are trained on much larger datasets. We have made our model and code public for reproducibility at https://microsoft.github.io/Magma.

Neural implicit scene representations have recently shown promising results in dense visual SLAM. However, existing implicit SLAM algorithms are constrained to single-agent scenarios, and fall difficulty in large indoor scenes and long sequences. Existing multi-agent SLAM frameworks cannot meet the constraints of communication bandwidth. To this end, we propose the first distributed multi-agent collaborative SLAM framework with distributed mapping and camera tracking, joint scene representation, intra-to-inter loop closure, and multi-submap fusion. Specifically, our proposed distributed neural mapping and tracking framework only needs peer-to-peer communication, which can greatly improve multi-agent cooperation and communication efficiency. A novel intra-to-inter loop closure method is designed to achieve local (single-agent) and global (multi-agent) consistency. Furthermore, to the best of our knowledge, there is no real-world dataset for NeRF-based/GS-based SLAM that provides both continuous-time trajectories groundtruth and high-accuracy 3D meshes groundtruth. To this end, we propose the first real-world indoor neural slam (INS) dataset covering both single-agent and multi-agent scenarios, ranging from small room to large-scale scenes, with high-accuracy ground truth for both 3D mesh and continuous-time camera trajectory. This dataset can advance the development of the community. Experiments on various datasets demonstrate the superiority of the proposed method in both mapping, tracking, and communication. The dataset and code will be open-source on https://github.com/dtc111111/MNESLAM.

Diffusion models have shown excellent performance in text-to-image generation. However, existing methods often suffer from performance bottlenecks when dealing with complex prompts involving multiple objects, characteristics, and relations. Therefore, we propose a Multi-agent Collaboration-based Compositional Diffusion (MCCD) for text-to-image generation for complex scenes. Specifically, we design a multi-agent collaboration based scene parsing module that generates an agent system containing multiple agents with different tasks using MLLMs to adequately extract multiple scene elements. In addition, Hierarchical Compositional diffusion utilizes Gaussian mask and filtering to achieve the refinement of bounding box regions and highlights objects through region enhancement for accurate and high-fidelity generation of complex scenes. Comprehensive experiments demonstrate that our MCCD significantly improves the performance of the baseline models in a training-free manner, which has a large advantage in complex scene generation. The code will be open-source on github.

Simultaneous localization and mapping (SLAM) systems with novel view synthesis capabilities are widely used in computer vision, with applications in augmented reality, robotics, and autonomous driving. However, existing approaches are limited to single-agent operation. Recent work has addressed this problem using a distributed neural scene representation. Unfortunately, existing methods are slow, cannot accurately render real-world data, are restricted to two agents, and have limited tracking accuracy. In contrast, we propose a rigidly deformable 3D Gaussian-based scene representation that dramatically speeds up the system. However, improving tracking accuracy and reconstructing a globally consistent map from multiple agents remains challenging due to trajectory drift and discrepancies across agents' observations. Therefore, we propose new tracking and map-merging mechanisms and integrate loop closure in the Gaussian-based SLAM pipeline. We evaluate \ours on synthetic and real-world datasets and find it more accurate and faster than the state of the art.

Real-time multi-agent collaboration for ego-motion estimation and high-fidelity 3D reconstruction is vital for scalable spatial intelligence. However, traditional methods produce sparse, low-detail maps, while recent dense mapping approaches struggle with high latency.To overcome these challenges, we present MAC-Ego3D, a novel framework for real-time collaborative photorealistic 3D reconstruction via Multi-Agent Gaussian Consensus. MAC-Ego3D enables agents to independently construct, align, and iteratively refine local maps using a unified Gaussian splat representation. Through Intra-Agent Gaussian Consensus, it enforces spatial coherence among neighboring Gaussian splats within an agent. For global alignment, parallelized Inter-Agent Gaussian Consensus, which asynchronously aligns and optimizes local maps by regularizing multi-agent Gaussian splats, seamlessly integrates them into a high-fidelity 3D model. Leveraging Gaussian primitives, MAC-Ego3D supports efficient RGB-D rendering, enabling rapid inter-agent Gaussian association and alignment.MAC-Ego3D bridges local precision and global coherence, delivering higher efficiency, largely reducing localization error, and improving mapping fidelity. It establishes a new SOTA on synthetic and real-world benchmarks, achieving a 15xincrease in inference speed, order-of-magnitude reductions in ego-motion estimation error for partial cases, and RGB PSNR gains of 4 to 10 dB. Our code will be made publicly available.