Transferring Hierarchical Structure with Dual Meta Imitation Learning

Chongkai Gao, Yizhou Jiang, Feng Chen

Keywords: imitation learning, meta learning, hierarchical structure, robot learning

Abstract: Hierarchical Imitation learning (HIL) is an effective way for robots to learn sub-skills from long-horizon unsegmented demonstrations. However, the learned hierarchical structure lacks the mechanism to transfer across multi-tasks or to new tasks, which makes them have to learn from scratch when facing a new situation. Transferring and reorganizing modular sub-skills require fast adaptation ability of the whole hierarchical structure. In this work, we propose Dual Meta Imitation Learning (DMIL), a hierarchical meta imitation learning method where the high-level network and sub-skills are iteratively meta-learned with model-agnostic meta-learning. DMIL uses the likelihood of state-action pairs from each sub-skill as the supervision for the high-level network adaptation, and use the adapted high-level network to determine different data set for each sub-skill adaptation. We theoretically prove the convergence of the iterative training process of DMIL and establish the connection between DMIL and the Expectation-Maximization algorithm. Empirically, we achieve state-of-the-art few-shot imitation learning performance on the meta-world benchmark.

One-sentence Summary: This paper presents dual meta imitation learning (DMIL) to meta learn a hierarchical network from multi-task demonstrations that can quickly adapt to new tasks with few shot demonstrations.