ASC me to Do Anything: Multi-task Training for Embodied AI

TL;DR

This paper introduces Atomic Skill Completion (ASC), a multi-task training approach for Embodied AI that leverages shared atomic skills to improve success rates and interpretability across multiple tasks.

Contribution

The paper proposes ASC, a novel pre-training scheme that decouples skill learning from high-level tasks, enabling effective multi-task training in Embodied AI environments.

Findings

ASC doubles success rates on seen scenes

ASC quadruples success rates on unseen scenes

Multi-task agents outperform independent single-task agents by 52%

Abstract

Embodied AI has seen steady progress across a diverse set of independent tasks. While these varied tasks have different end goals, the basic skills required to complete them successfully overlap significantly. In this paper, our goal is to leverage these shared skills to learn to perform multiple tasks jointly. We propose Atomic Skill Completion (ASC), an approach for multi-task training for Embodied AI, where a set of atomic skills shared across multiple tasks are composed together to perform the tasks. The key to the success of this approach is a pre-training scheme that decouples learning of the skills from the high-level tasks making joint training effective. We use ASC to train agents within the AI2-THOR environment to perform four interactive tasks jointly and find it to be remarkably effective. In a multi-task setting, ASC improves success rates by a factor of 2x on Seen scenes and 4x on Unseen scenes compared to no pre-training. Importantly, ASC enables us to train a multi-task agent that has a 52% higher Success Rate than training 4 independent single task agents. Finally, our hierarchical agents are more interpretable than traditional black-box architectures.