Contrastive Language, Action, and State Pre-training for Robot Learning

TL;DR

This paper presents CLASP, a novel pre-training method that unifies language, action, and state representations in robotics, enabling improved zero-shot retrieval, captioning, and reinforcement learning tasks through distributional learning.

Contribution

It introduces a distributional pre-training approach extending CLIP to better capture complex language-behaviour relationships in robotics.

Findings

Superior zero-shot retrieval and captioning performance on unseen datasets

Effective generation of meaningful exploratory behaviours from textual commands

Potential for broad generalization to various downstream robot learning tasks

Abstract

In this paper, we introduce a method for unifying language, action, and state information in a shared embedding space to facilitate a range of downstream tasks in robot learning. Our method, Contrastive Language, Action, and State Pre-training (CLASP), extends the CLIP formulation by incorporating distributional learning, capturing the inherent complexities and one-to-many relationships in behaviour-text alignment. By employing distributional outputs for both text and behaviour encoders, our model effectively associates diverse textual commands with a single behaviour and vice-versa. We demonstrate the utility of our method for the following downstream tasks: zero-shot text-behaviour retrieval, captioning unseen robot behaviours, and learning a behaviour prior for language-conditioned reinforcement learning. Our distributional encoders exhibit superior retrieval and captioning performance on unseen datasets, and the ability to generate meaningful exploratory behaviours from textual commands, capturing the intricate relationships between language, action, and state. This work represents an initial step towards developing a unified pre-trained model for robotics, with the potential to generalise to a broad range of downstream tasks.