Contrastive Language, Action, and State Pre-training for Robot Learning
This work addresses the challenge of integrating multimodal data for robotics, offering a foundational step towards a unified pre-trained model with potential for broad generalization, though it is incremental as it builds on CLIP.
The paper tackles the problem of unifying language, action, and state information for robot learning by introducing CLASP, a method that extends CLIP with distributional learning to handle one-to-many relationships, resulting in superior performance in zero-shot retrieval and captioning on unseen datasets and enabling meaningful exploratory behaviors from text.
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.