DexTrack: Towards Generalizable Neural Tracking Control for Dexterous Manipulation from Human References

TL;DR

DexTrack introduces a neural tracking controller trained on large-scale demonstrations, combining reinforcement and imitation learning, to enable generalizable dexterous manipulation in robots, demonstrating significant success rate improvements in simulation and real-world tests.

Contribution

The paper presents a novel data-driven approach that integrates homotopy optimization, reinforcement learning, and imitation learning to develop a generalizable neural controller for dexterous manipulation.

Findings

Over 10% success rate improvement over baselines.

Effective generalization in both simulation and real-world environments.

Enhanced demonstration diversity through homotopy optimization.

Abstract

We address the challenge of developing a generalizable neural tracking controller for dexterous manipulation from human references. This controller aims to manage a dexterous robot hand to manipulate diverse objects for various purposes defined by kinematic human-object interactions. Developing such a controller is complicated by the intricate contact dynamics of dexterous manipulation and the need for adaptivity, generalizability, and robustness. Current reinforcement learning and trajectory optimization methods often fall short due to their dependence on task-specific rewards or precise system models. We introduce an approach that curates large-scale successful robot tracking demonstrations, comprising pairs of human references and robot actions, to train a neural controller. Utilizing a data flywheel, we iteratively enhance the controller's performance, as well as the number and quality of successful tracking demonstrations. We exploit available tracking demonstrations and carefully integrate reinforcement learning and imitation learning to boost the controller's performance in dynamic environments. At the same time, to obtain high-quality tracking demonstrations, we individually optimize per-trajectory tracking by leveraging the learned tracking controller in a homotopy optimization method. The homotopy optimization, mimicking chain-of-thought, aids in solving challenging trajectory tracking problems to increase demonstration diversity. We showcase our success by training a generalizable neural controller and evaluating it in both simulation and real world. Our method achieves over a 10% improvement in success rates compared to leading baselines. The project website with animated results is available at https://meowuu7.github.io/DexTrack/.