Visuomotor Policy Learning

For robots to successfully transition from lab settings to everyday environments, they must begin to reason about the risks associated with their actions and make informed, risk-aware decisions. This is particularly true for robots performing mobile manipulation tasks, which involve both interacting with and navigating within dynamic, unstructured spaces. However, existing whole-body controllers for mobile manipulators typically lack explicit mechanisms for risk-sensitive decision-making under uncertainty. To our knowledge, we are the first to (i) learn risk-aware visuomotor policies for mobile manipulation conditioned on egocentric depth observations with runtime-adjustable risk sensitivity, and (ii) show risk-aware behaviours can be transferred through Imitation Learning (IL) to a visuomotor policy conditioned on egocentric depth observations. Our method achieves this by first training a privileged teacher policy using Distributional Reinforcement Learning (DRL), with a risk-neutral distributional critic. Distortion risk-metrics are then applied to the critic's predicted return distribution to calculate risk-adjusted advantage estimates used in policy updates to achieve a range of risk-aware behaviours. We then distil teacher policies with IL to obtain risk-aware student policies conditioned on egocentric depth observations. We perform extensive evaluations demonstrating that our trained visuomotor policies exhibit risk-aware behaviour (specifically achieving better worst-case performance) while performing reactive whole-body motions in unmapped environments, leveraging live depth observations for perception.

Visuomotor navigation policies have shown strong perception-action coupling for embodied agents, yet they often struggle with safe navigation and dynamic obstacle avoidance in complex real-world environments. We introduce CHOP, a novel approach that leverages Counterfactual Human Preference Labels to align visuomotor navigation policies towards human intuition of safety and obstacle avoidance in navigation. In CHOP, for each visual observation, the robot's executed trajectory is included among a set of counterfactual navigation trajectories: alternative trajectories the robot could have followed under identical conditions. Human annotators provide pairwise preference labels over these trajectories based on anticipated outcomes such as collision risk and path efficiency. These aggregated preferences are then used to fine-tune visuomotor navigation policies, aligning their behavior with human preferences in navigation. Experiments on the SCAND dataset show that visuomotor navigation policies fine-tuned with CHOP reduce near-collision events by 49.7%, decrease deviation from human-preferred trajectories by 45.0%, and increase average obstacle clearance by 19.8% on average across multiple state-of-the-art models, compared to their pretrained baselines. These improvements transfer to real-world deployments on a Ghost Robotics Vision60 quadruped, where CHOP-aligned policies improve average goal success rates by 24.4%, increase minimum obstacle clearance by 6.8%, reduce collision and intervention events by 45.7%, and improve normalized path completion by 38.6% on average across navigation scenarios, compared to their pretrained baselines. Our results highlight the value of counterfactual preference supervision in bridging the gap between large-scale visuomotor policies and human-aligned, safety-aware embodied navigation.

Despite the rapid progress of vision-language-action (VLA) models, the prevailing practice of predicting action chunks as discrete waypoints remains structurally misaligned with the intrinsic continuity of physical motion. This discretization arises naturally from fixed-rate robot data collection and the token-by-token prediction paradigm of large language models, but ties actions to rigid sampling rates, does not naturally support analytically consistent higher-order derivatives, and introduces quantization artifacts that hinder precise, compliant interaction. We propose Neural Implicit Action Fields (NIAF), which reformulates chunk-level action representation from discrete waypoints to continuous action functions. Using a vision-language model as a hierarchical spectral modulator over a learnable motion prior, NIAF synthesizes continuous-time action manifolds with arbitrary temporal resolution. This formulation enables analytical differentiation, allowing explicit supervision of velocity and regularization of higher-order derivative signals to promote mathematical consistency, physical plausibility, and control smoothness. Our approach achieves strong results on CALVIN and LIBERO across diverse backbones. Real-world experiments further confirm that NIAF supports stable impedance control, bridging policy-side action generation and execution-side smooth control.

Humans rehearse possible futures offline, as in mental practice and perhaps dreaming, suggesting that world models may support task learning away from the environment. Standard machine learning world models compress visual input into latent vectors, discarding the spatial structure that characterizes sensory cortex. We propose isomorphic world models: architectures that preserve sensory topology, so physics prediction becomes geometric propagation rather than abstract state transition. We implement this idea with motor-gated neural fields, where activity evolves through local lateral connectivity and motor commands multiplicatively modulate specific channels. Across three experiments, the same architecture learns ballistic prediction without ``teleporting,'' improves a catching policy offline by propagating task error through a frozen learned world model, and develops body-selective motor channels without body labels. These results provide preliminary evidence that physical prediction, offline task learning, and body-linked representation share a common computational substrate: action-conditional prediction within a spatial map.

Contact-rich teleoperation logs expose a policy I/O design choice: demonstrations may contain the robot execution (E), leader command (C), or both. These signals are not interchangeable: E2E may discard contact-generating command offsets, whereas E2C preserves these offsets but omits the robot's execution response. We propose Dual-State Conditioning (EC2C), which conditions on both E and C while predicting future C, exposing command-execution mismatch as a cue for contact, latency, payload, and operator compensation; in quasi-static contact, this cue is often force-correlated. On a low-cost setup without force, tactile, or motor-current policy input, EC2C outperforms E2E and a strong E2C baseline across several real-world contact-rich, force-sensitive, and dynamic tasks. These results support EC2C as a practical default I/O setting for contact-rich imitation learning. We further formulate latency-adaptive inpainting as a temporal extension of this I/O choice for action-chunking policies, and discuss when long histories help dynamic inference or introduce causal confounding.

Vision-language-action (VLA) models typically inject proprioception only as a late conditioning signal, preventing robot state from grounding instruction understanding or directing visual attention. We introduce ThinkProprio, which discretizes proprioception into VLM-vocabulary tokens and uses them jointly with the instruction to gate visual patches before VLM computation, steering the model toward action-relevant evidence while discarding redundant tokens early. We find that proprioception added as a passive conditioning signal leaves performance essentially unchanged; its value emerges when token-form state acts as an active query that, with the instruction, selects which visual patches the VLM processes. Systematic ablations show that VLM-vocabulary tokens outperform learned projectors as the state encoding, and that retaining only about \SI{12}{\percent} of the visual tokens surpasses on CALVIN ABC$\to$D. Across CALVIN, LIBERO, and real-world manipulation, ThinkProprio reduces end-to-end inference latency while improving the matched full-token baseline.

Robot manipulation uses temporal context to select actions and visual foresight to assess their consequences, yet dense representations of past and future observations incur substantial processing costs. We introduce PACT-WAM, a world-action model that jointly generates a 16-step action trajectory and its temporally corresponding visual forecast through conditional flow sampling. Hierarchical history encoding assigns coarse spatial representations to earlier observations and finer representations to recent ones, retaining 16 observations with 256 tokens per view, 75% fewer than dense encoding of the same frames. A shared flow module jointly updates continuous action and visual states through two modality-specific heads under transition-wise causal attention, and a TiTok-VAE decoder reconstructs multi-view future images from the visual latents. Decoded forecasts also support Proposal Review (PR), a vision-language model component for execution-prefix selection and proposal rejection. Without PR, PACT-WAM achieves average success rates of 98.6%, 92.3%, and 78.0% on LIBERO, RoboTwin 2.0, and real-world Piper tasks, respectively. PR provides a test-time enhancement, raising these rates to 99.5%, 93.4%, and 86.7%. Ablations show that hierarchical history allocation and joint action-visual generation improve control success, while analyses of visual capacity and forecast-guided execution characterize the trade-offs between success and proposal-generation cost.

Vision-based imitation learning has enabled impressive robotic manipulation skills, but action imitation alone provides limited supervision of the geometric consequences of robot behavior. To address this limitation, we introduce Implicit Scene Supervision (ISS) Policy, a 3D visuomotor diffusion policy with a DiT backbone that predicts continuous action sequences from point-cloud observations. ISS augments action diffusion with a supervised robot motion predictor that maps generated actions and robot-state context to end-effector motion, and then uses the predicted motion together with gripper intent to forecast future point-cloud representations. By explicitly modeling the intermediate transition from action to robot motion, ISS encourages the policy to capture how its actions affect the surrounding 3D scene. We further introduce asymmetric gradient routing to separate direct motion regression from scene-level policy supervision, together with a change-balanced objective that accounts for variations in scene-change magnitude. These auxiliary objectives provide dynamics-aware geometric supervision using only expert demonstrations, without requiring additional annotations or auxiliary modules at inference time. ISS Policy achieves state-of-the-art performance on single-arm manipulation tasks in MetaWorld and dexterous manipulation tasks in Adroit, while real-world dual-arm experiments further demonstrate its effectiveness on physical robotic manipulation. The resulting framework preserves the scalable DiT backbone and standard diffusion-policy control interface. Code and videos will be released.

Generalist vision--language--action (VLA) policies are typically trained on heterogeneous mixtures of robot demonstrations spanning diverse embodiments, action spaces, and observation configurations. Modeling such heterogeneity with a shared dense action module can induce negative transfer, particularly when action spaces or visual observations differ across data sources. We address this issue with HiMoE-VLA, a VLA framework built around a Hierarchical Mixture-of-Experts (HiMoE) action module. HiMoE uses Action-Space MoE layers at the input/output boundaries to specialize computation for distinct action spaces, Heterogeneity-Balancing MoE layers in neighboring layers to provide balanced capacity for residual variation in observations, scenes, and embodiments, and dense Transformer blocks in the middle to integrate shared representations. Two auxiliary objectives further guide this hierarchy: a contrastive Action-Space Regularization objective for boundary specialization and a load-balancing objective for stable expert utilization. HiMoE-VLA reaches 3.98 on CALVIN, 98.0% on LIBERO, and 75.0% and 63.7% average success on real xArm7 and ALOHA tasks; under controlled heterogeneous co-training, it turns the negative transfer observed in strong baselines into positive transfer. The code and models are publicly available at https://github.com/ZhiyingDu/HiMoE-VLA.

Developing efficient and accurate visuomotor policies poses a central challenge in robotic imitation learning. While recent rectified flow approaches have advanced visuomotor policy learning, they suffer from a key limitation: After iterative distillation, generated actions may deviate from the ground-truth actions corresponding to the current visual observation, leading to accumulated error as the reflow process repeats and unstable task execution. We present Selective Flow Alignment (SeFA), an efficient and accurate visuomotor policy learning framework. SeFA resolves this challenge by a selective flow alignment strategy, which leverages expert demonstrations to selectively correct generated actions and restore consistency with observations, while preserving multimodality. This design introduces a consistency correction mechanism that ensures generated actions remain observation-aligned without sacrificing the efficiency of one-step flow inference. Extensive experiments across both simulated and real-world manipulation tasks show that SeFA Policy surpasses state-of-the-art diffusion-based and flow-based policies, achieving superior accuracy and robustness while reducing inference latency by over 98%. By unifying rectified flow efficiency with observation-consistent action generation, SeFA provides a scalable and dependable solution for real-time visuomotor policy learning. Code is available on https://github.com/RongXueZoe/SeFA.

Generative BC policies model expert action distributions well, but they do not explicitly reason about what scene a candidate action would produce, so they cannot tell whether it reaches a task-relevant state. We address this by coupling action generation with an action-conditioned dynamics model that predicts the future observations induced by actions. However, naively placing such a dynamics model inside a generative policy creates a circular dependency. Future observations are needed to guide action generation, yet those observations can only be predicted from the action samples being generated. We introduce Dynamics-Aligned Flow Matching Policy (DAP), where a flow matching policy and an action-conditioned flow matching dynamics model co-sample actions and future observations through flow extrapolation. At each step the two flows correct each other, and the samples converge to a self-consistent action and future pair on which the policy and dynamics model agree. Across four simulated tasks and three real-world tasks, DAP achieves the highest average success rate, with relative improvements of 56.5% in simulation and 33.3% in the real world over the strongest baseline in each setting. Moreover, when the future-observation estimate is corrupted at inference time, DAP recovers a self-consistent action-future pair through iterative refinement, whereas video-conditioned and post-hoc dynamics baselines, which use a fixed future surrogate, degrade or fail.

Likelihood-free inference (LFI) enables system identification in complex tasks via black-box modelling, abstracting nonlinearity and stochasticity, and infers a domain distribution for adapting agents to parametric deployment conditions. LFI assumes an arbitrary support for sampling, which remains fixed as the initial generic prior is refined to increasingly descriptive posteriors. Misspecified support can therefore yield suboptimal yet overconfident posteriors. We address this issue by using the posterior of an inference step to guide the adaptation of the support using three illustrative heuristics: EDGE, MODE, and CENTRE. Each heuristic interprets the updated belief and enables support adaptation alongside posterior inference. For illustrative purposes, we first study misspecified support in LFI and evaluate the utility of our heuristics using stochastic dynamical benchmarks. We then evaluate posterior-driven heuristic support adaptation for parameter inference and policy learning in a dynamic deformable linear object (DLO) manipulation task. Inference results in a finer length and stiffness classification for a parametric set of DLOs. When the resulting posteriors are used as domain distributions for sim-based policy learning, they lead to more robust object-centric agent performance.

Geometry-grounded vision models, such as VGGT, have emerged as robust visual encoders, providing essential geometric priors for robotic manipulation. However, the high computational cost of these models often leads to slow inference, limiting their practical applications in real-world robotics. This paper introduces eVGGT, a lightweight geometry-aware vision encoder distilled from the high-performing VGGT. Our findings demonstrate two primary advantages: i) integrating eVGGT into imitation learning frameworks (including ACT and Diffusion Policy) yields up to a 6.3% improvement in success rate over standard 2D encoders across bimanual and single-arm tasks in both simulation and real-world settings with variable viewpoints; ii) eVGGT achieves a nearly 5 times speedup and a 63% reduction in memory usage compared to state-of-the-art geometry-aware encoders while maintaining comparable task performance. These results suggest that eVGGT substantially alleviates the performance-latency bottleneck that has limited geometry-aware visuomotor policies in real-world deployment.

Visuomotor policies trained via behavior cloning are vulnerable to covariate shift, where small deviations from expert trajectories can compound into failure. Common strategies to mitigate this issue involve expanding the training distribution through human-in-the-loop corrections or synthetic data augmentation. However, these approaches are often labor-intensive, rely on strong task assumptions, or compromise the quality of imitation. We introduce Latent Policy Barrier, a framework for robust visuomotor policy learning. Inspired by Control Barrier Functions, LPB treats the latent embeddings of expert demonstrations as an implicit barrier separating safe, in-distribution states from unsafe, out-of-distribution (OOD) ones. Our approach decouples the role of precise expert imitation and OOD recovery into two separate modules: a base diffusion policy solely on expert data, and a dynamics model trained on both expert and suboptimal policy rollout data. At inference time, the dynamics model predicts future latent states and optimizes them to stay within the expert distribution. Both simulated and real-world experiments show that LPB improves both policy robustness and data efficiency, enabling reliable manipulation from limited expert data and without additional human correction or annotation.

Human vision is a highly active process driven by gaze, which directs attention to task-relevant regions through foveation, dramatically reducing visual processing. In contrast, robot learning systems typically rely on passive, uniform processing of raw camera images. In this work, we explore how incorporating human-like active gaze into robotic policies can enhance efficiency and robustness. We develop GIAVA (Gaze Integrated Active-Vision ALOHA), a robot vision system that emulates human head and neck movement, and gaze adjustment for foveated processing. Extending the AV-ALOHA robot platform, we introduce a framework for simultaneously collecting eye-tracking, perspective control, and robot manipulation demonstration data from a human operator. We also open-source a simulation benchmark and dataset for training robot policies that incorporate human gaze. Inspired by recent work in foveated image segmentation and given the widespread use of Vision Transformers (ViTs) in robot learning, we integrate gaze information into ViTs using a foveated patch tokenization scheme. Compared to uniform patch tokenization, this significantly reduces the number of tokens, and thus computation. Our results show that our method for foveated robot vision drastically reduces computational overhead, and enhances robustness to background distractors. Notably, on certain high-precision tasks, foveated vision also improves performance, as reflected in higher success rates. Together, these findings suggest that human-inspired foveated visual processing offers untapped potential and should be further considered as a useful inductive bias in robotic vision systems. https://soltanilara.github.io/giava/

Conventional visuomotor imitation learning usually predicts future robot actions directly in the time domain. Such formulations often have limited physical scene awareness and weak memory. In this work, we propose Wavelet Policy, a lightweight imitation learning framework that combines World Prior Memory (WPM) with wavelet-based multi-scale action modeling. Our key idea is to encode persistent physical scene structure from static background images into compact memory tokens, which are fused into world-prior tokens and injected into the encoder during forward propagation. Based on this memory-conditioned representation, we further perform wavelet-domain decomposition over horizon-aligned latent action tokens and adopt a Single-Encoder Multiple-Decoder (SE2MD) architecture to model latent components at different temporal scales. The resulting latent subbands are reconstructed through inverse wavelet transform and finally projected into executable action chunks. To facilitate efficient world prior learning, we introduce a world-prior adaptation loss, encouraging the background encoder to retain persistent scene knowledge while remaining lightweight and stable. Extensive experiments on four simulated and six real-world robotic manipulation tasks show that Wavelet Policy consistently outperforms strong baselines. These results demonstrate that combining scale-domain action modeling with world-prior memory provides an effective and efficient solution for embodied manipulation.