Robot Imitation Learning

Contact-rich manipulation requires robots to balance accurate motion tracking with compliant interaction, yet most visual-action policies leave compliance fixed at the controller level. We present Imp-ACT, a methodologically grounded and practical approach to incorporating direction-dependent Cartesian stiffness modulation directly into demonstration collection, without manual stiffness selection or offline target reconstruction. During teleoperation, a self-tuning impedance controller adapts stiffness along the instantaneous direction of motion while maintaining compliance in orthogonal directions. The adapted stiffness is applied and recorded alongside visual observations and motion commands, capturing motion and compliance under the same dynamics. We implement this pipeline using Action Chunking with Transformer (ACT) to predict end-effector pose, gripper action, and motion-direction stiffness from visual, proprioceptive, and wrench observations. The performance of Imp-ACT is evaluated on wiping and plug insertion using both success rate and quantitative measures of contact behavior. Compared with fixed low- and high-stiffness baselines, Imp-ACT achieves comparable or higher success while maintaining low interaction forces. In wiping, it reduces contact-force vibration by approximately $29\times$ relative to the compliant baseline and $180\times$ relative to the stiff baseline. In plug insertion, it reduces forces orthogonal to the insertion direction by $43\%$ relative to the better fixed-stiffness baseline. These results highlight the benefit of maintaining sufficient stiffness along the direction needed for task execution while preserving compliance in other directions to limit contact forces and accommodate environmental constraints.

Achieving robust cross-embodiment generalization in imitation learning demands overcoming a critical representation flaw that inextricably entangles task semantics with hardware-specific visual geometry. We propose an interaction-centric framework that leverages the shared structure of two-finger grippers via a parameterized universal gripper abstraction, yielding a canonical gripper-frame representation. Given language and RGB-D observations, a VLM infers the subtask and grounds an interaction triplet (gripper, held, target), while SAM~2.1 tracks masks to reduce VLM queries. We design concise hybrid features that combine target/collision artificial potential fields for global guidance with segmented gripper-frame point clouds for local geometry, and use a Flow-Matching Transformer to predict smooth 7-DoF action chunks. Experiments in simulation and real-world tasks demonstrate that ours is the first imitation learning approach to simultaneously achieve competitive benchmark scores and extreme cross-embodiment/cross-viewpoint zero-shot sim-to-real transfer to completely distinct, heterogeneous robot platforms.

Simulation enables scalable training of Vision-Language-Action policies by using privileged experts to generate visual demonstrations without requiring every trajectory to be collected through manual teleoperation. However, such pipelines typically retain successful demonstrations while failed rollouts are discarded, even though they expose precisely the off-nominal states from which recovery must be learned. We introduce Kintsugi-VLA, a framework for converting failed rollouts into targeted synthetic recovery data by exploiting exact state restoration and branching in simulation. For a fixed privileged expert, we define interventional recoverability as the probability of completing the original task after the simulator is restored to a given state, estimate it using adaptive Monte Carlo continuations with pointwise Wilson intervals, and characterize its non-monotonic evolution along failed trajectories. These estimates identify an observed terminal low-recoverability frontier-the point after which measured recoverability remains below a threshold-which is then used to select informative recovery starting states. In a simulated Franka manipulation task, targeted recovery data yield aggregate SmolVLA recovery success of 34.6% and 38.4% under difficulty- and frame-budget matching, respectively, 5.8 and 6.7 percentage points above uniform sampling within the same recovery window. The same ordering is observed under disturbed end-to-end execution and shifted clutter and physics conditions, while clean-task success decreases from 76.8% to 74.7%. Kintsugi-VLA demonstrates how failed simulator rollouts can be transformed from discarded experience into structured recovery-training data through direct interventional measurement.

Vision-Language-Action models achieve strong performance in robot manipulation, but often require large numbers of parameters. In this work, we propose the Future Representation Action Model (FRAM), a small policy that explicitly links the future end-effector trajectory to the current visual input. FRAM uses the image coordinates of the predicted trajectory as spatial pointers and reads local visual features related to the motion from the current image. This organizes the information for action generation into the reference position (Where), the visual state (What), and the future motion (Future). Trajectory labels are generated automatically from demonstrations and camera geometry, so no manual annotation is needed. With 138.7M parameters, including a frozen language encoder, FRAM reaches an average success rate of 92.2% over the four standard LIBERO suites, close to the 94.2% of $π_0$ with 3.3B parameters. Without extra training, it also reaches an average of 67.3% on LIBERO-Plus. Ablations confirm that both the future trajectory and the local visual features improve performance and robustness. On a real dual-arm UR5e, FRAM stacks cups using only wrist cameras, including choosing and switching between the left and right arms. These results show that selecting visual information based on future motion is an effective way to obtain both high performance and robustness in a small robot policy.

Human demonstrations are a cheap source of data for dexterous manipulation, but co-training a robot policy on them requires closing the human--robot gap in every modality the policy consumes. We present VisTacAlign, a framework for co-training 3D-visual-tactile dexterous policies on human and robot demonstrations. Glove-tracked human hand motion is retargeted to a 17-DoF tactile robot hand with a one-time fingertip correction. The human hand is then erased from both stereo views and replaced by a posed robot-hand mesh painted with pixels from robot recordings, and a real-time stereo foundation model is re-run on the composite, so the human point clouds carry the same stereo errors and visibility as the robot ones. Finally, a capacitive tactile glove is aligned to the robot's fingertip sensors in its signal space, giving one interpretable per-finger force representation. A diffusion transformer consumes point-cloud, proprioceptive, and per-finger tactile tokens. On three real-world tasks requiring precise force -- Lego assembly, plucking strawberries of varying size, and activating and lifting a power drill -- adding aligned human demonstrations to existing robot data improves over robot-only policies, and ablations show that both tactile input and visual alignment are necessary. Project page: https://vis-tac-align.github.io

Contact-rich assembly tasks such as peg-in-hole insertion remain difficult to learn from limited demonstrations. While retrieval-augmented imitation learning, which augments target demonstrations with relevant prior data, offers a promising direction, its applicability to contact-rich manipulation remains largely unexplored. Contact-rich insertion unfolds over multiple phases from search to insert, and retrieving phase-specific experience from prior data in principled ways remains an open question. Our key insight is that a compliant wrist enables the robot to sustain contact throughout execution, producing rich tactile and force signals that naturally reveal the phase structure of insertion and inform what should be retrieved. Based on this insight, we present PHASE (PHase-Aware Segmentation and REtrieval), a framework for compliance-enabled tactile phase retrieval that integrates multimodal contact-aware representation learning, variable-length phase segmentation from tactile signals, and phase-consistent retrieval for policy learning. We evaluate PHASE on real-world peg-in-hole insertion across five peg geometries, comparing against retrieval strategies drawn from state-of-the-art methods under a shared policy architecture. PHASE improves the overall success rate by 13 percentage points over the strongest non-phase-aware baseline, and improves performance under unseen initial positions by 30 percentage points. These results demonstrate that aligning retrieval with interaction-defined contact phases substantially improves robustness in few-shot insertion learning.

Mobile humanoid manipulation requires both reaching a usable workspace and preserving precise hand-object interactions as object poses and contact conditions change. Learning these behaviors from limited task-specific data remains challenging. To bridge this gap, we introduce Praxis, a whole-body manipulation framework that combines physical interaction priors from one-shot egocentric video demonstrations with closed-loop posture calibration and online perception. The framework coordinates three stages: vision-language-guided navigation toward target objects, closed-loop posture calibration to align the arm-hand workspace, and dexterous manipulation with synchronized upper- and lower-body control. Online visual feedback re-grounds demonstrated interaction geometry under new object poses and scene configurations, while tactile feedback adapts hand motions to actual contact conditions. Each manipulation skill is specified by one human demonstration, without task-specific manipulation-policy retraining. Experiments across five long-horizon manipulation tasks demonstrate spatial, visual, and cross-object generalization, as well as recovery from external physical disturbances across all three stages.

For people who are blind, touch provides an essen-tial channel for accessing written information through Braille. Bringing a similar capability to robots requires them not only to recognize tactile patterns, but also to actively establish physical contact that makes those patterns readable. Yet existing robotic Braille readers largely focus on recognition after contact, leaving contact establishment itself insufficiently addressed. We present an adaptive-contact framework for robotic tactile Braille reading that assesses contact quality and physically corrects unsuitable contact before recognition and reconstruc-tion. Multi-Head Policy Learning uses expert-guided contact-adjustment demonstrations to jointly learn contact acceptability and pose corrections. During deployment, the robot iteratively evaluates and re-establishes contact, retaining reliable tactile observations for pose-aware fusion and Braille reconstruction. Across 20 physical Braille plates used for learning and eval-uation, the proposed approach achieves 94.0% tactile quality and 88.6% tactile reconstruction on the ten online-evaluation plates. These results demonstrate the importance of actively establishing readable contact, rather than relying solely on recognition under imperfect tactile observations, for reliable robotic Braille reading.

Policies trained with imitation learning can accumulate errors over time, causing the robot to drift outside the training distribution. Existing methods mitigate this covariate shift by collecting additional data where the policy fails or is likely to fail. The first places the robot in unsafe conditions and the second requires choosing an appropriate noise distribution to collect new expert demonstrations under that noise. We propose Policy-Calibrated DAgger, a method that makes use of the properties of recent generative policies to estimate the policy's noise offline by using its own predicted action distribution. We measure a diffusion policy's spread of predicted actions at observations along the expert trajectory and measure its closed-loop error relative to a recorded trajectory. To address issues with measuring error in a multimodal action space, we guide the policy towards the trajectory during closed-loop control through partial denoising, and use properties of a diffusion model to unnormalize the measured error as if we did not guide it. We experiment in a scenario where a robot is tasked to reach an engine lever in a cluttered and narrow environment and show results in a 3D photorealistic simulator and a 2D planar reacher environment. We show that our method surpasses policies trained with dataset aggregation without noising and matches the performance of the best noise level in hindsight, without requiring a sweep over noise levels.

We present POIL, a point-based one-shot imitation learning framework with stable dynamical systems. While one-shot imitation avoids collecting extensive demonstrations, successful one-shot manipulation requires not only transferring a demonstrated trajectory to a novel object but also executing it robustly under changing scene conditions, grasp configurations, and external disturbances. POIL addresses both problems through a shared representation: a set of 3D points on the object's functional part, used jointly for trajectory transfer and closed-loop execution. The one-shot transfer from the demonstrated trajectory is enabled with point correspondences. POIL grounds the shared functional part with a multi-modal large language model, and transfers the trajectory across viewpoint, pose, and object category changes. During execution, multi-view tracking observes the same points online, and Point-set BCSDM drives them in closed loop by projecting per-point velocities onto a single rigid-body twist computed from the tracked points alone. This extends stable dynamical models from an SE(3) pose to a point set without requiring a known 3D model or pose estimator. We show that at the goal the controller becomes a gradient flow on the classical SO(3) potential, so its terminal phase inherits the almost-global convergence of that potential under a rigid-object assumption. Across simulation and real-robot experiments, POIL transfers a single demonstration across object category, grasp pose, and goal geometry, while recovering from external disturbances during execution. Project page: https://sangminkim-99.github.io/poil

Imitation learning enables robots to acquire manipulation skills from demonstrations, but the resulting policies can fail outside the training data, while collecting more demonstrations requires substantial human effort. Human-in-the-loop reinforcement learning uses corrective feedback during online training, but typically learns the complete task policy rather than refining a pretrained imitation policy. We introduce Res-HIL, a human-in-the-loop residual reinforcement learning framework that learns corrective actions on top of a frozen imitation policy. Each human intervention provides two complementary learning signals: direct supervision of the residual policy and reward shaping of preceding autonomous behavior. Res-HIL combines these signals with zero initialization of the residual policy to stabilize and accelerate online learning. We evaluate Res-HIL on five contact-rich manipulation tasks spanning high-precision and long-horizon behaviors. With only 20 initial demonstrations, Res-HIL outperforms state-of-the-art full-policy human-in-the-loop reinforcement learning and residual fine-tuning without human guidance on every task after ten minutes of online training. Res-HIL improves its pretrained base policies and outperforms imitation policies trained with five times more demonstrations. An ablation study shows that direct residual supervision is critical to performance, while intervention-aware reward shaping substantially improves training efficiency.

General-purpose vision-language models (VLMs) bring broad knowledge and spatial reasoning to robot manipulation, yet existing systems either use them indirectly, to predict constraints or write programs, or give them a view of the scene rather than a world in which to act. We present World Action Agent (WAA), a multi-agent harness through which VLMs pilot robots with basic tools, making every decision within a visual action workspace. The workspace has three properties. Contact views, selected automatically from the scene geometry, present the scene around the current interaction. Action rehearsal turns each action into an editable proposal that the agent, alone or through an Imagination Agent, previews and revises against planning feedback before execution. In-view correction closes the loop between observation, rehearsal, and low-level execution, letting the agent remove residual offsets in the view where it observes them. Through the same workspace, WAA acquires embodied procedural knowledge in two ways: it evolves multimodal skills from expert videos and human teaching under evidence-based review and consults them through a Skill Agent, and its interaction traces train smaller VLMs to pilot the same harness. On LIBERO-Pro, WAA with skills evolved only from LIBERO-90 reaches a state-of-the-art 75.6% average success, outperforming end-to-end VLAs, code-as-policy agents, and a visual-harness baseline with the same backbone; the same skills remain effective on robosuite without further learning. Fine-tuning Qwen3.5-9B on harness traces raises its out-of-domain success from 1.7% to 43.3%.

Robotic imitation learning often relies on external cameras, yet local interaction cues such as object proximity, contact onset, and grasp state are difficult to observe near the fingertips because of occlusion and limited temporal resolution. We study how to effectively incorporate complementary fingertip sensing into imitation learning using pressure-sensitive tactile and reflective proximity sensors, along with pretrained sensor encoders. The two modalities provide information at different manipulation phases: proximity sensing is informative before contact, whereas tactile sensing becomes informative after contact. However, naively adding these signals to a policy does not consistently improve performance and can even underperform vision-only policies, suggesting that sparse, phase-dependent sensor signals are difficult to exploit from limited demonstrations. We therefore propose a proprioception-anchored pretraining method, PROprioceptive-and-PRoactive Anchoring (PROPRA), which independently aligns each fingertip sensor history with proprioceptive and action segments. This provides a continuously available sensorimotor reference, allowing each sensor to be aligned independently during its informative phases. Experiments on real-world manipulation tasks show that our pretraining method improves average success rates over vision-only policies and image-anchored pretraining baselines. Representation analysis further shows that it preserves richer information about pre-contact states, enabling more effective use of complementary fingertip sensing. Please refer to our project page: https://tomohiromotoda.github.io/nia.propra/

Humans typically rely on vision, touch, hearing, and proprioception to perceive contact and adapt their actions during manipulation. Providing robots with comparable responsiveness therefore requires hardware that can retain and use these complementary sensory signals. Most imitation-learning systems, however, observe demonstrations primarily through vision and proprioception, limiting access to contact information that is difficult to infer visually. We present PolyUMI, an open-source platform for scalable visual--tactile--audio demonstration collection and robot deployment. Its lightweight, wireless handheld gripper records synchronized wrist-camera, optical tactile, contact-audio, and proprioceptive observations without requiring a tethered workstation. The same sensing finger can be transferred to the robot end effector, preserving the sensing geometry between demonstration collection and policy execution. To effectively use these heterogeneous observations, we further introduce VisTA, a token-level multimodal policy that integrates information across sensors and time to predict contact-aware robot actions. Experiments spanning object inference, slip control, and contact-rich manipulation show that touch and audio reveal task-relevant information beyond vision and that VisTA is competitive with or outperforms existing multimodal policies. Together, PolyUMI and VisTA provide an accessible pipeline for collecting multimodal demonstrations and learning policies that perceive physical interaction beyond vision. Project Page: https://polyumi-vista.github.io

Robot manipulation policies trained through imitation learning inherit not only the demonstrated behavior but also the conservative execution tempo of robot demonstrations. Existing acceleration approaches can execute faster than the original demonstrations, but determine the appropriate acceleration primarily from robot-side information or a predefined set of tempo factors, leaving open how to obtain a task-appropriate reference for how fast each manipulation phase should progress. We introduce EgoSpeedUp, a framework that uses human manipulation as temporal supervision for robot imitation learning. Our key insight is that human demonstrations naturally reveal task-appropriate, phase-wise manipulation tempo. Given slow robot demonstrations and human demonstrations of the same task, EgoSpeedUp aligns corresponding manipulation phases, estimates their relative execution tempos from multiple human demonstrations, and transfers the resulting phase-wise tempo by retiming the robot demonstrations. The retimed demonstrations are then used for standard behavior cloning, allowing the robot to retain its executable manipulation behavior while learning to perform it at a human-informed tempo. Across two real-world manipulation tasks, EgoSpeedUp improves the task success rate by an average of 25 percentage points (pp) while reducing successful execution time by 36.5%. These results demonstrate that human manipulation tempo provides an effective temporal reference for learning faster and more reliable robot policies.

Skilled humans can catch fast-moving objects softly by coordinating interception, velocity matching, and follow-through to mitigate impact. Learning such impact-aware catching with reinforcement learning (RL), however, is challenging, as the policy must achieve reliable interception and grasping while regulating the sensitive transition into contact. Moreover, even a capable privileged-state RL teacher may not provide ideal demonstrations for a deployable imitation-learning (IL) student: teacher failures limit task coverage, while small variations in pre-contact motion can produce substantially different impact and grasping outcomes. We characterize this phenomenon through interventional outcome sensitivity and introduce the outcome-sensitive window (OSW) to guide targeted demonstration construction. Building on this formulation, we propose Outcome-Sensitive Motion Search, which learns a task-conditioned manifold of successful OSW motions and performs local geodesic search to refine successful teacher rollouts and repair task conditions where the teacher fails. We then validate candidate motions through complete rollouts under a calibrated IL-student action-error model and retain only successful executions as demonstrations. Extensive simulation experiments demonstrate that our method effectively repairs task conditions where the teacher fails and enables the resulting IL policy to outperform the privileged RL teacher in both catching success and impact mitigation.

Most robots manipulate objects solely with their end effectors, whereas humans flexibly leverage different body parts, such as the forearm and elbow, especially when handling oversized objects. Learning such whole-arm manipulation is chal-lenging due to long-horizon sparse rewards, limited contact sens-ing, and the sim-to-real gap in contact and actuator dynamics. To address these challenges, we propose Current-Aligned Link Manipulation, a framework for learning long-horizon contact-rich manipulation using motor current as joint load related feedback. Three stage-specific policies first learn repositioning, grasping, and lifting using privileged simulation information, and a stage router sequences them to generate complete task demonstrations. For sim-to-real transfer, a causal current mapper predicts physical motor current from simulated joint histories, aligning the actuator current observation between simulation and hardware. A unified student policy then learns from these demonstrations using only deployable sensor observations and is further refined with DAgger. The task policies are trained entirely in simulation, and the final student is deployed on hardware. Experiments demonstrate 76.2% (762/1000 trials) complete-task success in simulation and 73.3% success (22/30 trials) on the physical robot for sequential oversized-object lifting.

Generalization in robotic manipulation requires policies to perform tasks across diverse unseen object instances that vary in shape, size, and pose. However, conventional behavior cloning (BC) methods often overfit to instance-specific geometry and appearance, limiting transfer to novel objects. We introduce KeyGen, a framework that learns canonicalized semantic 3D keypoints from point clouds and uses them as structured object-centric representations for policy learning. A visuomotor diffusion policy conditions on these keypoints together with object-centric geometry to predict full manipulation trajectories, enabling consistent geometric correspondence across object instances. To evaluate category-level generalization, we construct a photorealistic simulation benchmark with three manipulation tasks and a planning-driven data generation pipeline that produces expert trajectories across diverse object instances. Experiments show that KeyGen significantly outperforms prior methods on both seen and unseen objects under pose variation, scales effectively with additional demonstrations per object, maintains robustness to object rescaling, and achieves strong performance in both simulation and real-world manipulation.

Human hand-object interactions (HOIs) provide a rich source of demonstrations for dexterous manipulation, but learning directly from them presents challenges in bridging morphology gaps, ensuring dynamical feasibility, and sim-to-real deployment. We present Morphometric Imitation, a three-stage framework that transforms reconstructed HOIs into zero-shot sim-to-real visuomotor policies. First, morphometric optimization (MMO) kinematically retargets human motion across hand morphologies while preserving demonstrated contacts. Second, residual reinforcement learning (RL) refines the kinematic reference using object pose and contact information from the human motion to produce dynamically feasible robot demonstrations. Third, these demonstrations are distilled into visuomotor policies. Across three robot hands and ten HOIs, MMO outperforms five baselines in contact F1, improving on the strongest ones by 8 to 28 points, while improving the success rate of downstream dynamic retargeting by as much as 35 points. On a Sharpa hand, the visuomotor policies achieve 89.3% zero-shot success in 300 real-world trials on 30 objects spanning 10 categories. Project page: https://morphometricimitation.github.io

An always-on robot faces an endless stream that never resets: instructions arrive and lapse, the scene changes, and its own past actions reshape what it must reason about. Today's action models are built for the opposite: a fixed instruction, no mid-task intervention, single-step reasoning. In an open-ended world a robot must watch a live stream for far-future cues, recall its own far-past actions, and act on them under dual-arm concurrency. We present ARMS (Always-on Robot in Multi-modal Streams), a deliberately simple streaming policy: a single pretrained $π$0.5 backbone augmented by three lightweight modules that turn live perception, embodied states, and the robot's own past actions into context the backbone reads before it acts. The modules update this context asynchronously, so watching and recalling never block acting and the two arms act at once. Rather than inventing new mechanisms, ARMS integrates these learned context providers with an agent-causal self-history that logs which arm did what, and when. To supervise them without extra annotation, we build ARMS Dataset, whose staged construction script itself labels every module from real dual-arm teleoperation. Trained on it, ARMS reaches 45% on the combined task against 28% for the strongest of our four main baselines, and ablations confirm the memory module, the embodied-state head, and asynchronous concurrency are each necessary.

Advances in generative modeling have recently been extensively employed in robotics for policy learning. In particular, Conditional Flow Matching (CFM) trained with expert demonstrations has been shown to outperform existing methods on robot manipulation benchmarks. While prior work has mainly focused on single-task settings, we study the problem from a multi-task perspective, as training independent models for each task is computationally expensive. Multi-Task policy learning comes with its own set of challenges, as naively training on a concatenated dataset of demonstrations would either require increased model capacity to accommodate the added complexity or result in drops in performance. We propose to distill knowledge from single-task CFM experts into a shared multi-task policy by transferring their learned velocity fields. We combine this distillation signal with the original CFM objective to retain fidelity to the demonstrations. Experiments on RLBench show that our approach improves multi-task policy performance over naive training while maintaining a fixed model size.

Modern robot imitation learning increasingly relies on generative policies based on diffusion or flow-matching models, which generate actions by transforming samples from a prior distribution. A key question is whether the choice of prior matters. Replacing the standard Gaussian with a closer-to-target, non-Gaussian prior has been shown to substantially improve performance when training from scratch. A natural next step is to ask whether these gains transfer to fine-tuning pretrained Large Behavior Models (LBMs) such as LBM 1.0, $π_{0.5}$, and GR00T~N1.5, where one might expect even larger gains. Surprisingly, we find that this is not the case, except possibly at very low fine-tuning data fractions. Across over 100K simulation rollouts spanning all three aforementioned LBMs on 40+ tasks in two simulation platforms, and 1250 hardware rollouts on five bimanual manipulation tasks, non-Gaussian priors that are demonstrably closer to the target yield statistically indistinguishable or worse fine-tuning performance than a standard Gaussian prior. Diagnostic analyses suggest why: fine-tuned imitation learning policies converge to similar action predictions across priors, despite their fine-tuned encoder embeddings diverging substantially from the pretrained embeddings and each other. A learning-rate ablation further confirms that encoder training is the dominant factor in fine-tuning performance, substantially outweighing the effect of prior choice. We conclude with concrete directions for future research on when and why learned priors might still matter in fine-tuning. Project page: https://cxu-tri.github.io/non_gaussian_FT/

Runway walking requires coordinated control of posture, stride, foot placement, and whole-body motion to effectively present clothing and convey a distinctive style. However, humanoid robots used in fashion shows typically rely on locomotion policies optimized primarily for stability and walking speed, limiting their ability to reproduce expressive, human-like runway motions. In this work, we present an end-to-end framework that transforms monocular runway videos into deployable humanoid locomotion policies through motion recovery, robot retargeting, motion correction, policy training, simulation-based evaluation, and physical deployment. We evaluate the proposed framework on the Booster K1 humanoid robot using runway-style catwalk motions. The learned policy completed every physical trial without falling, while reproducing the characteristic narrow foot placement and coordinated movement of the legs, torso, and arms. The results demonstrate that our proposed training framework enables the Booster K1 to perform stable and expressive catwalk motions, highlighting its potential for humanoid robotic applications in fashion shows and other performance-oriented scenarios.

Musculoskeletal (MSK) humanoids provide a physiologically grounded embodiment for studying full-body motor control, but their high-dimensional muscle actuation, delayed activation dynamics, and redundant muscle--tendon structures make learning substantially harder than torque-driven humanoid control. Existing MSK benchmarks remain fragmented across gait, prosthetics, dexterous hands, or challenge-specific tracks, leaving full-body muscle-actuated control insufficiently evaluated under standardized tasks, methods, and metrics. We introduce MSK-Bench, a benchmark of 22 full-body motor-control tasks organized into three progressively challenging categories: postural stabilization, common locomotor behaviors, and contact-rich environmental interaction. Under unified task protocols and robustness perturbations, MSK-Bench evaluates 5 representative control paradigms, including reward-based RL, agentic reward tuning, latent-action RL, imitation-prior control, and residual adaptation over imitation priors. Beyond task success and reward, MSK-Bench further reports robustness analysis and physiology-oriented diagnostics, including activation cost, joint smoothness, and EMG-envelope similarity. Our empirical study shows that embodiment-aware exploration and structured action representations improve task coverage in high-dimensional muscle spaces, imitation priors enhance reference-compatible stabilization and locomotion but degrade under contact-rich terrain mismatch, and residual adaptation can recover successful behaviors when fixed references fail. We further find that improved task success does not necessarily imply improved physiological agreement, highlighting the importance of evaluating task performance, robustness, and physiological behavior jointly. MSK-Bench provides a task--method--metric testbed for full-body muscle-actuated humanoid control.

Humanoid robots require diverse embodied experiences to acquire complex loco-manipulation and collaborative skills. However, existing humanoid data pipelines primarily focus on individual agents, while physical multi-robot collaboration remains difficult to scale due to costly hardware, dedicated spaces, and repeated resets. In this work, we introduce MATE, a Multi-Agent virtual TEleoperation platform for humanoid collaboration data collection that enables multiple geographically distributed operators to simultaneously control whole-body humanoids in a shared physics-based environment. MATE removes the need for multiple physical robots and co-located operation while preserving physically coupled interactions among humanoids, objects, and environments. Using MATE, we construct a multi-humanoid collaboration dataset comprising 24.1 hours of coordinated behavior across 2,500 joint episodes and five long-horizon tasks, including object handover, relay delivery, environment interaction, and cooperative transport. To improve learning from these interaction-rich demonstrations, we introduce EAIS, an Execution-Aligned Interaction Sampling strategy that computes sampling signals within an execution-aligned prefix and prioritizes task-progressing and interaction-critical behaviors. We evaluate MATE with representative imitation learning and vision-language-action policies across diverse collaboration tasks. Experiments demonstrate efficient data collection, effective policy learning, and zero-shot transfer from virtual demonstrations to a physical humanoid without real-world fine-tuning. Project page: https://yerik-yu.github.io/MATE/

How should a robot learn to manipulate objects so fragile that sub-Newton contact forces can cause irreversible damage? Existing visuo-tactile policy learning typically treats tactile sensing as an additional policy input. In direct-contact force-sensitive manipulation, however, the bottleneck can arise earlier, during data collection: manual gripper control is too delayed and coarse-grained to reliably maintain the narrow force range required for stable grasping. We therefore use a deterministic 25 Hz tactile reflex controller as a collection-time teacher, producing demonstrations with controller-shaped grasping behavior for tactile-free policy learning. On Action Chunking with Transformers (ACT), policies trained from reflex-shaped demonstrations recover the teacher's grasping profile and achieve 95% stable grasps on the nominal plastic-cup task, substantially outperforming visually screened manual demonstrations. The same intervention improves in-distribution stability on $π_{0.5}$ and shows a favorable exploratory trend on an unseen paper-cup variant. Under randomized external disturbance, however, the reflex-data $π_{0.5}$ policy still fails in 45% of policy-only trials, whereas a deployment-time reflex arbiter retains all grasps. These results reveal a new role for tactile feedback in force-sensitive manipulation: rather than integrating tactile into the policy, we use it as a collection-time teacher that shapes grasping behavior in demonstrations for policy learning, while disturbance rejection remains controller-dependent, revealing the boundary of tactile-free policy.

Humanoid motion tracking policies rely on dense frame-by-frame references, limiting their use as high-level motion controllers for planning and interactive motion generation. We study Sparse Timed Keyframe Motion Tracking, where a policy receives only sparse future keyframes and their desired arrival times, and must execute stable whole-body motions that reach successive goals. We propose PLAT, a three-stage sparse timed keyframe motion tracking policy learning framework with Privileged LAtent Transition learning. PLAT bridges dense motion tracking and sparse goal-conditioned control by exploiting dense goal sequences as privileged supervision during training while requiring only sparse timed keyframe commands at deployment. A pretrained dense tracking expert first provides robust motion priors. A privileged latent prior is then learned through DAgger-style imitation, followed by latent residual reinforcement learning that refines latent transitions instead of directly optimizing actions. Extensive simulation experiments demonstrate that PLAT maintains accurate and stable sparse timed keyframe tracking across varying planning horizons, with particularly strong performance under long-horizon commands. Successful deployment on a Unitree G1 humanoid robot further demonstrates the effectiveness and practicality of PLAT for sparse humanoid motion control.

Most minimally invasive surgery is still performed with hand-held laparoscopic instruments, and the surgeon's instrument kinematics are lost when the operation ends; only the endoscope video is kept. This paper presents an end-to-end pipeline that captures this motion in the operating room and uses it to train a surgical robot policy, validated on live animals. We introduce a surgical instrument-state logger that mounts on the shaft of a standard laparoscopic instrument and recovers its pose and jaw state from an inertial sensor, a time-of-flight sensor and a Hall sensor, with no external camera or tracker. A data pipeline measures the latency of every sensor channel against a robot ground truth and aligns the channels before forming observation-action pairs. On these demonstrations we train a diffusion policy with a fine-tuned DINOv3 backbone, selecting its design by closed-loop rollouts in a physics simulator reconstructed from depth maps of an ex-vivo rabbit appendix. The policy is then retrained on 849 in-vivo demonstrations from four live rabbits and deployed on four additional live rabbits with electrosurgery armed. With the surgeon selecting the surgical phase, the policy completed the appendectomy in three of the four animals. The results show that demonstrations recorded from a surgeon's own instruments are sufficient to train, select and deploy a bimanual surgical policy in vivo. The robot serves only as the timing reference for sensor calibration and as the executor, and collects no demonstrations. Both demonstration corpora are released to support future surgical robot learning research.

Behavior cloning for robot manipulation relies on expert demonstrations. However, for tasks that require dynamic stability, precise contact timing, or dexterous coordination, human operators may find it hard or even impossible to collect data. We study this infeasible-demonstration regime and propose GLIDE: Guardrails for Learning from Infeasible Demonstrations Efficiently, a framework that infers task-specific failure modes and converts them into executable guardrails for data collection and policy deployment. Given a task description and the conditioning teleoperation code, GLIDE writes guardrails that use system states to filter teleoperation and policy commands, constrain failure-prone actions, and iteratively improve from trajectory feedback. Across three tasks, GLIDE discovers emergent guardrails that go beyond domain-expert hardcoded ones, improving data collection over naive VR teleoperation and domain-expert hardcoded guardrails. After refinement, GLIDE raises data-collection success from 0-10 percent to 70-90 percent across the three tasks. During policy execution, mixed-data guarded policies reach 70 percent, 60 percent, and 60 percent success on Tomato plate transfer, Marker handover and stand, and Wine serving tasks. These results show that GLIDE can support policy learning when direct demonstrations are infeasible. Project website: http://guardrail-policy.github.io/

Learning robot manipulation policies from human video demonstrations constitutes a promising avenue for scalable robot learning. However, comparing different human-to-robot (H2R) transfer methods remains challenging, as existing approaches are evaluated under different settings, including differing task suites, scene layouts, object instances, and amounts of robot supervision. To address this challenge, we present H2RBench, a Real2Sim benchmark for evaluating H2R transfer methods. H2RBench provides a standardized protocol built on real human video demonstrations and simulated robot demonstrations, and includes four manipulation tasks spanning diverse interaction requirements. We evaluate multiple representative H2R transfer methods, each adopting a different strategy for bridging the embodiment gap. Using H2RBench, we systematically characterize how each method scales with the amount of human demonstrations, revealing that methods differ substantially in their ability to leverage additional human data. We further show that simulation performance is broadly predictive of real-world robot performance, with an overall Pearson correlation of r = 0.89, Spearman correlation of \r{ho} = 0.85 and Mean Maximum Rank Violation (MMRV) of 0.06 across method-task configurations. These results establish H2RBench as a practical and scalable benchmark for comparative H2R evaluation prior to real-world deployment.