Imitation learning frameworks for robotic manipulation have drawn attention in the recent development of language model grounded robotics. However, the success of the frameworks largely depends on the coverage of the demonstration cases: When the demonstration set does not include examples of how to act in all possible situations, the action may fail and can result in cascading errors. To solve this problem, we propose a framework that uses serialized Finite State Machine (FSM) to generate demonstrations and improve the success rate in manipulation tasks requiring a long sequence of precise interactions. To validate its effectiveness, we use environmentally evolving and long-horizon puzzles that require long sequential actions. Experimental results show that our approach achieves a success rate of up to 98 in these tasks, compared to the controlled condition using existing approaches, which only had a success rate of up to 60, and, in some tasks, almost failed completely.
Mobile robots are essential in applications such as autonomous delivery and hospitality services. Applying learning-based methods to address mobile robot tasks has gained popularity due to its robustness and generalizability. Traditional methods such as Imitation Learning (IL) and Reinforcement Learning (RL) offer adaptability but require large datasets, carefully crafted reward functions, and face sim-to-real gaps, making them challenging for efficient and safe real-world deployment. We propose an online human-in-the-loop learning method PVP4Real that combines IL and RL to address these issues. PVP4Real enables efficient real-time policy learning from online human intervention and demonstration, without reward or any pretraining, significantly improving data efficiency and training safety. We validate our method by training two different robots -- a legged quadruped, and a wheeled delivery robot -- in two mobile robot tasks, one of which even uses raw RGBD image as observation. The training finishes within 15 minutes. Our experiments show the promising future of human-in-the-loop learning in addressing the data efficiency issue in real-world robotic tasks. More information is available at: https://metadriverse.github.io/pvp4real/
Imitation Learning can train robots to perform complex and diverse manipulation tasks, but learned policies are brittle with observations outside of the training distribution. 3D scene representations that incorporate observations from calibrated RGBD cameras have been proposed as a way to mitigate this, but in our evaluations with unseen embodiments and camera viewpoints they show only modest improvement. To address those challenges, we propose Adapt3R, a general-purpose 3D observation encoder which synthesizes data from calibrated RGBD cameras into a vector that can be used as conditioning for arbitrary IL algorithms. The key idea is to use a pretrained 2D backbone to extract semantic information, using 3D only as a medium to localize this information with respect to the end-effector. We show across 93 simulated and 6 real tasks that when trained end-to-end with a variety of IL algorithms, Adapt3R maintains these algorithms' learning capacity while enabling zero-shot transfer to novel embodiments and camera poses.
Autonomous agents and systems cover a number of application areas, from robotics and digital assistants to combinatorial optimization, all sharing common, unresolved research challenges. It is not sufficient for agents to merely solve a given task; they must generalize to out-of-distribution tasks, perform reliably, and use hardware resources efficiently during training and inference, among other requirements. Several methods, such as reinforcement learning and imitation learning, are commonly used to tackle these problems, each with different trade-offs. However, there is a lack of benchmarking suites that define the environments, datasets, and metrics which can be used to provide a meaningful way for the community to compare progress on applying these methods to real-world problems. We introduce A2Perf--a benchmark with three environments that closely resemble real-world domains: computer chip floorplanning, web navigation, and quadruped locomotion. A2Perf provides metrics that track task performance, generalization, system resource efficiency, and reliability, which are all critical to real-world applications. Using A2Perf, we demonstrate that web navigation agents can achieve latencies comparable to human reaction times on consumer hardware, reveal reliability trade-offs between algorithms for quadruped locomotion, and quantify the energy costs of different learning approaches for computer chip-design. In addition, we propose a data cost metric to account for the cost incurred acquiring offline data for imitation learning and hybrid algorithms, which allows us to better compare these approaches. A2Perf also contains several standard baselines, enabling apples-to-apples comparisons across methods and facilitating progress in real-world autonomy. As an open-source benchmark, A2Perf is designed to remain accessible, up-to-date, and useful to the research community over the long term.
Many robot demonstration datasets contain heterogeneous demonstrations of varying quality. This heterogeneity may benefit policy pre-training, but can hinder robot performance when used with a final imitation learning objective. In particular, some strategies in the data may be less reliable than others or may be underrepresented in the data, leading to poor performance when such strategies are sampled at test time. Moreover, such unreliable or underrepresented strategies can be difficult even for people to discern, and sifting through demonstration datasets is time-consuming and costly. On the other hand, policy performance when trained on such demonstrations can reflect the reliability of different strategies. We thus propose for robots to self-curate based on online robot experience (Demo-SCORE). More specifically, we train and cross-validate a classifier to discern successful policy roll-outs from unsuccessful ones and use the classifier to filter heterogeneous demonstration datasets. Our experiments in simulation and the real world show that Demo-SCORE can effectively identify suboptimal demonstrations without manual curation. Notably, Demo-SCORE achieves over 15-35% higher absolute success rate in the resulting policy compared to the base policy trained with all original demonstrations.
This paper presents ArticuBot, in which a single learned policy enables a robotics system to open diverse categories of unseen articulated objects in the real world. This task has long been challenging for robotics due to the large variations in the geometry, size, and articulation types of such objects. Our system, Articubot, consists of three parts: generating a large number of demonstrations in physics-based simulation, distilling all generated demonstrations into a point cloud-based neural policy via imitation learning, and performing zero-shot sim2real transfer to real robotics systems. Utilizing sampling-based grasping and motion planning, our demonstration generalization pipeline is fast and effective, generating a total of 42.3k demonstrations over 322 training articulated objects. For policy learning, we propose a novel hierarchical policy representation, in which the high-level policy learns the sub-goal for the end-effector, and the low-level policy learns how to move the end-effector conditioned on the predicted goal. We demonstrate that this hierarchical approach achieves much better object-level generalization compared to the non-hierarchical version. We further propose a novel weighted displacement model for the high-level policy that grounds the prediction into the existing 3D structure of the scene, outperforming alternative policy representations. We show that our learned policy can zero-shot transfer to three different real robot settings: a fixed table-top Franka arm across two different labs, and an X-Arm on a mobile base, opening multiple unseen articulated objects across two labs, real lounges, and kitchens. Videos and code can be found on our project website: https://articubot.github.io/.
Humans can accomplish complex contact-rich tasks using vision and touch, with highly reactive capabilities such as quick adjustments to environmental changes and adaptive control of contact forces; however, this remains challenging for robots. Existing visual imitation learning (IL) approaches rely on action chunking to model complex behaviors, which lacks the ability to respond instantly to real-time tactile feedback during the chunk execution. Furthermore, most teleoperation systems struggle to provide fine-grained tactile / force feedback, which limits the range of tasks that can be performed. To address these challenges, we introduce TactAR, a low-cost teleoperation system that provides real-time tactile feedback through Augmented Reality (AR), along with Reactive Diffusion Policy (RDP), a novel slow-fast visual-tactile imitation learning algorithm for learning contact-rich manipulation skills. RDP employs a two-level hierarchy: (1) a slow latent diffusion policy for predicting high-level action chunks in latent space at low frequency, (2) a fast asymmetric tokenizer for closed-loop tactile feedback control at high frequency. This design enables both complex trajectory modeling and quick reactive behavior within a unified framework. Through extensive evaluation across three challenging contact-rich tasks, RDP significantly improves performance compared to state-of-the-art visual IL baselines through rapid response to tactile / force feedback. Furthermore, experiments show that RDP is applicable across different tactile / force sensors. Code and videos are available on https://reactive-diffusion-policy.github.io.
In this paper, we introduce CAPS (Context-Aware Priority Sampling), a novel method designed to enhance data efficiency in learning-based autonomous driving systems. CAPS addresses the challenge of imbalanced training datasets in imitation learning by leveraging Vector Quantized Variational Autoencoders (VQ-VAEs). The use of VQ-VAE provides a structured and interpretable data representation, which helps reveal meaningful patterns in the data. These patterns are used to group the data into clusters, with each sample being assigned a cluster ID. The cluster IDs are then used to re-balance the dataset, ensuring that rare yet valuable samples receive higher priority during training. By ensuring a more diverse and informative training set, CAPS improves the generalization of the trained planner across a wide range of driving scenarios. We evaluate our method through closed-loop simulations in the CARLA environment. The results on Bench2Drive scenarios demonstrate that our framework outperforms state-of-the-art methods, leading to notable improvements in model performance.
Diffusion models have revolutionized imitation learning, allowing robots to replicate complex behaviours. However, diffusion often relies on cameras and other exteroceptive sensors to observe the environment and lacks long-term memory. In space, military, and underwater applications, robots must be highly robust to failures in exteroceptive sensors, operating using only proprioceptive information. In this paper, we propose ProDapt, a method of incorporating long-term memory of previous contacts between the robot and the environment in the diffusion process, allowing it to complete tasks using only proprioceptive data. This is achieved by identifying "keypoints", essential past observations maintained as inputs to the policy. We test our approach using a UR10e robotic arm in both simulation and real experiments and demonstrate the necessity of this long-term memory for task completion.
End-to-end vision-based imitation learning has demonstrated promising results in autonomous driving by learning control commands directly from expert demonstrations. However, traditional approaches rely on either regressionbased models, which provide precise control but lack confidence estimation, or classification-based models, which offer confidence scores but suffer from reduced precision due to discretization. This limitation makes it challenging to quantify the reliability of predicted actions and apply corrections when necessary. In this work, we introduce a dual-head neural network architecture that integrates both regression and classification heads to improve decision reliability in imitation learning. The regression head predicts continuous driving actions, while the classification head estimates confidence, enabling a correction mechanism that adjusts actions in low-confidence scenarios, enhancing driving stability. We evaluate our approach in a closed-loop setting within the CARLA simulator, demonstrating its ability to detect uncertain actions, estimate confidence, and apply real-time corrections. Experimental results show that our method reduces lane deviation and improves trajectory accuracy by up to 50%, outperforming conventional regression-only models. These findings highlight the potential of classification-guided confidence estimation in enhancing the robustness of vision-based imitation learning for autonomous driving. The source code is available at https://github.com/ElaheDlv/Confidence_Aware_IL.
In recent research advancements within the community, large language models (LLMs) have sparked great interest in creating autonomous agents. However, current prompt-based agents often heavily rely on large-scale LLMs. Meanwhile, although fine-tuning methods significantly enhance the capabilities of smaller LLMs, the fine-tuned agents often lack the potential for self-reflection and self-improvement. To address these challenges, we introduce a novel agent framework named RetroAct, which is a framework that jointly optimizes both task-planning and self-reflective evolution capabilities in language agents. Specifically, we develop a two-stage joint optimization process that integrates imitation learning and reinforcement learning, and design an off-policy joint policy gradient optimization algorithm with imitation learning regularization to enhance the data efficiency and training stability in agent tasks. RetroAct significantly improves the performance of open-source models, reduces dependency on closed-source LLMs, and enables fine-tuned agents to learn and evolve continuously. We conduct extensive experiments across various testing environments, demonstrating RetroAct has substantial improvements in task performance and decision-making processes.
This paper presents a novel Perceptual Motor Learning (PML) framework integrated with Active Inference (AIF) to enhance lateral control in Highly Automated Vehicles (HAVs). PML, inspired by human motor learning, emphasizes the seamless integration of perception and action, enabling efficient decision-making in dynamic environments. Traditional autonomous driving approaches--including modular pipelines, imitation learning, and reinforcement learning--struggle with adaptability, generalization, and computational efficiency. In contrast, PML with AIF leverages a generative model to minimize prediction error ("surprise") and actively shape vehicle control based on learned perceptual-motor representations. Our approach unifies deep learning with active inference principles, allowing HAVs to perform lane-keeping maneuvers with minimal data and without extensive retraining across different environments. Extensive experiments in the CARLA simulator demonstrate that PML with AIF enhances adaptability without increasing computational overhead while achieving performance comparable to conventional methods. These findings highlight the potential of PML-driven active inference as a robust alternative for real-world autonomous driving applications.
Dexterous grasping remains a fundamental yet challenging problem in robotics. A general-purpose robot must be capable of grasping diverse objects in arbitrary scenarios. However, existing research typically relies on restrictive assumptions, such as single-object settings or limited environments, leading to constrained generalization. We present DexGraspVLA, a hierarchical framework for general dexterous grasping in cluttered scenes based on RGB image perception and language instructions. It utilizes a pre-trained Vision-Language model as the high-level task planner and learns a diffusion-based policy as the low-level Action controller. The key insight to achieve robust generalization lies in iteratively transforming diverse language and visual inputs into domain-invariant representations via foundation models, where imitation learning can be effectively applied due to the alleviation of domain shift. Notably, our method achieves a 90+% success rate under thousands of unseen object, lighting, and background combinations in a "zero-shot" environment. Empirical analysis confirms the consistency of internal model behavior across environmental variations, thereby validating our design and explaining its generalization performance. DexGraspVLA also demonstrates free-form long-horizon prompt execution, robustness to adversarial objects and human disturbance, and failure recovery, which are rarely achieved simultaneously in prior work. Extended application to nonprehensile object grasping further proves its generality. Code, model, and video are available at dexgraspvla.github.io.
We present the POPGym Arcade, a collection of hardware-accelerated, pixel-based environments with shared observation and action spaces. Each environment includes fully and partially observable variants, enabling counterfactual studies on partial observability. We also introduce mathematical tools for analyzing policies under partial observability, which reveal how agents recall past information to make decisions. Our analysis shows (1) that controlling for partial observability is critical and (2) that agents with long-term memory learn brittle policies that struggle to generalize. Finally, we demonstrate that recurrent policies can be "poisoned" by old, out-of-distribution observations, with implications for sim-to-real transfer, imitation learning, and offline reinforcement learning.
This paper proposes FABG (Facial Affective Behavior Generation), an end-to-end imitation learning system for human-robot interaction, designed to generate natural and fluid facial affective behaviors. In interaction, effectively obtaining high-quality demonstrations remains a challenge. In this work, we develop an immersive virtual reality (VR) demonstration system that allows operators to perceive stereoscopic environments. This system ensures "the operator's visual perception matches the robot's sensory input" and "the operator's actions directly determine the robot's behaviors" - as if the operator replaces the robot in human interaction engagements. We propose a prediction-driven latency compensation strategy to reduce robotic reaction delays and enhance interaction fluency. FABG naturally acquires human interactive behaviors and subconscious motions driven by intuition, eliminating manual behavior scripting. We deploy FABG on a real-world 25-degree-of-freedom (DoF) humanoid robot, validating its effectiveness through four fundamental interaction tasks: expression response, dynamic gaze, foveated attention, and gesture recognition, supported by data collection and policy training. Project website: https://cybergenies.github.io
This paper introduces the SOAR framework for imitation learning. SOAR is an algorithmic template that learns a policy from expert demonstrations with a primal dual style algorithm that alternates cost and policy updates. Within the policy updates, the SOAR framework uses an actor critic method with multiple critics to estimate the critic uncertainty and build an optimistic critic fundamental to drive exploration. When instantiated in the tabular setting, we get a provable algorithm with guarantees that matches the best known results in $\epsilon$. Practically, the SOAR template is shown to boost consistently the performance of imitation learning algorithms based on Soft Actor Critic such as f-IRL, ML-IRL and CSIL in several MuJoCo environments. Overall, thanks to SOAR, the required number of episodes to achieve the same performance is reduced by half.
We introduce a novel Inverse Reinforcement Learning (IRL) approach that overcomes limitations of fixed reward assignments and constrained flexibility in implicit reward regularization. By extending the Maximum Entropy IRL framework with a squared temporal-difference (TD) regularizer and adaptive targets, dynamically adjusted during training, our method indirectly optimizes a reward function while incorporating reinforcement learning principles. Furthermore, we integrate distributional RL to capture richer return information. Our approach achieves state-of-the-art performance on challenging MuJoCo tasks, demonstrating expert-level results on the Humanoid task with only 3 demonstrations. Extensive experiments and ablation studies validate the effectiveness of our method, providing insights into adaptive targets and reward dynamics in imitation learning.
Recent vision-language-action models (VLAs) build upon pretrained vision-language models and leverage diverse robot datasets to demonstrate strong task execution, language following ability, and semantic generalization. Despite these successes, VLAs struggle with novel robot setups and require fine-tuning to achieve good performance, yet how to most effectively fine-tune them is unclear given many possible strategies. In this work, we study key VLA adaptation design choices such as different action decoding schemes, action representations, and learning objectives for fine-tuning, using OpenVLA as our representative base model. Our empirical analysis informs an Optimized Fine-Tuning (OFT) recipe that integrates parallel decoding, action chunking, a continuous action representation, and a simple L1 regression-based learning objective to altogether improve inference efficiency, policy performance, and flexibility in the model's input-output specifications. We propose OpenVLA-OFT, an instantiation of this recipe, which sets a new state of the art on the LIBERO simulation benchmark, significantly boosting OpenVLA's average success rate across four task suites from 76.5% to 97.1% while increasing action generation throughput by 26$\times$. In real-world evaluations, our fine-tuning recipe enables OpenVLA to successfully execute dexterous, high-frequency control tasks on a bimanual ALOHA robot and outperform other VLAs ($\pi_0$ and RDT-1B) fine-tuned using their default recipes, as well as strong imitation learning policies trained from scratch (Diffusion Policy and ACT) by up to 15% (absolute) in average success rate. We release code for OFT and pretrained model checkpoints at https://openvla-oft.github.io/.
Imitation learning has proven to be highly effective in teaching robots dexterous manipulation skills. However, it typically relies on large amounts of human demonstration data, which limits its scalability and applicability in dynamic, real-world environments. One key challenge in this context is object generalization, where a robot trained to perform a task with one object, such as "hand over the apple," struggles to transfer its skills to a semantically similar but visually different object, such as "hand over the peach." This gap in generalization to new objects beyond those in the same category has yet to be adequately addressed in previous work on end-to-end visuomotor policy learning. In this paper, we present a simple yet effective approach for achieving object generalization through Vision-Language-Action (VLA) models, referred to as \textbf{ObjectVLA}. Our model enables robots to generalize learned skills to novel objects without requiring explicit human demonstrations for each new target object. By leveraging vision-language pair data, our method provides a lightweight and scalable way to inject knowledge about the target object, establishing an implicit link between the object and the desired action. We evaluate ObjectVLA on a real robotic platform, demonstrating its ability to generalize across 100 novel objects with a 64\% success rate in selecting objects not seen during training. Furthermore, we propose a more accessible method for enhancing object generalization in VLA models, using a smartphone to capture a few images and fine-tune the pre-trained model. These results highlight the effectiveness of our approach in enabling object-level generalization and reducing the need for extensive human demonstrations, paving the way for more flexible and scalable robotic learning systems.
Personal AI assistants (e.g., Apple Intelligence, Meta AI) offer proactive recommendations that simplify everyday tasks, but their reliance on sensitive user data raises concerns about privacy and trust. To address these challenges, we introduce the Guardian of Data (GOD), a secure, privacy-preserving framework for training and evaluating AI assistants directly on-device. Unlike traditional benchmarks, the GOD model measures how well assistants can anticipate user needs-such as suggesting gifts-while protecting user data and autonomy. Functioning like an AI school, it addresses the cold start problem by simulating user queries and employing a curriculum-based approach to refine the performance of each assistant. Running within a Trusted Execution Environment (TEE), it safeguards user data while applying reinforcement and imitation learning to refine AI recommendations. A token-based incentive system encourages users to share data securely, creating a data flywheel that drives continuous improvement. Specifically, users mine with their data, and the mining rate is determined by GOD's evaluation of how well their AI assistant understands them across categories such as shopping, social interactions, productivity, trading, and Web3. By integrating privacy, personalization, and trust, the GOD model provides a scalable, responsible path for advancing personal AI assistants. For community collaboration, part of the framework is open-sourced at https://github.com/PIN-AI/God-Model.