REVIEW 6 cited by
Teach a Robot to FISH: Versatile Imitation from One Minute of Demonstrations
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
While imitation learning provides us with an efficient toolkit to train robots, learning skills that are robust to environment variations remains a significant challenge. Current approaches address this challenge by relying either on large amounts of demonstrations that span environment variations or on handcrafted reward functions that require state estimates. Both directions are not scalable to fast imitation. In this work, we present Fast Imitation of Skills from Humans (FISH), a new imitation learning approach that can learn robust visual skills with less than a minute of human demonstrations. Given a weak base-policy trained by offline imitation of demonstrations, FISH computes rewards that correspond to the "match" between the robot's behavior and the demonstrations. These rewards are then used to adaptively update a residual policy that adds on to the base-policy. Across all tasks, FISH requires at most twenty minutes of interactive learning to imitate demonstrations on object configurations that were not seen in the demonstrations. Importantly, FISH is constructed to be versatile, which allows it to be used across robot morphologies (e.g. xArm, Allegro, Stretch) and camera configurations (e.g. third-person, eye-in-hand). Our experimental evaluations on 9 different tasks show that FISH achieves an average success rate of 93%, which is around 3.8x higher than prior state-of-the-art methods.
Forward citations
Cited by 6 Pith papers
-
OmniTacTune: Policy-Agnostic Real-World RL for Tactile Residual Adaptation of Visual Policies
A policy-agnostic two-stage real-world RL method learns tactile residual corrections on frozen visual policies, lifting contact-rich task success from 5–40% to 85–100% in under 80 minutes.
-
CDP: Towards Robust Autoregressive Visuomotor Policy Learning via Causal Diffusion
Causal Diffusion Policy adds historical action conditioning and attention cache sharing to diffusion-based robot policies, improving success rates on most tested manipulation tasks under degraded observations.
-
Residual Reward Models for Preference-based Reinforcement Learning
Combining a hand-designed or learned prior reward with a preference-trained residual improves sample efficiency and final performance in preference-based reinforcement learning.
-
3DFlowAction: Learning Cross-Embodiment Manipulation from 3D Flow World Model
A diffusion world model predicts 3D optical flow as an embodiment-agnostic action plan, and constrained optimization converts the flow into robot arm actions.
-
Is Optimal Transport Necessary for Inverse Reinforcement Learning?
Simple nearest-neighbor and segment-matching reward functions match or beat Optimal Transport based Inverse RL across 32 benchmarks.
-
Spatial-Temporal Aware Visuomotor Diffusion Policy Learning
A diffusion-based visuomotor policy gains 3D and 4D scene awareness from a dynamic Gaussian world model, improving simulated and real robot manipulation success rates.
Discussion (0). Continue with ORCID to comment.