Learning Adaptive Cross-Embodiment Visuomotor Policy with Contrastive Prompt Orchestration

TL;DR

This paper introduces CAPO, a novel contrastive prompt learning method with adaptive orchestration, enabling embodied agents to adapt efficiently to diverse and unseen environments by dynamically focusing on relevant domain factors.

Contribution

The paper proposes a hybrid contrastive learning strategy and an adaptive prompt orchestration mechanism for improved cross-embodiment visuomotor policy adaptation.

Findings

Outperforms state-of-the-art baselines in sample efficiency and performance

Demonstrates superior zero-shot adaptation to unseen environments

Effectively isolates task-relevant features from domain variations

Abstract

Learning adaptive visuomotor policies for embodied agents remains a formidable challenge, particularly when facing cross-embodiment variations such as diverse sensor configurations and dynamic properties. Conventional learning approaches often struggle to separate task-relevant features from domain-specific variations (e.g., lighting, field-of-view, and rotation), leading to poor sample efficiency and catastrophic failure in unseen environments. To bridge this gap, we propose ContrAstive Prompt Orchestration (CAPO), a novel approach for learning visuomotor policies that integrates contrastive prompt learning and adaptive prompt orchestration. For prompt learning, we devise a hybrid contrastive learning strategy that integrates visual, temporal action, and text objectives to establish a pool of learnable prompts, where each prompt induces a visual representation encapsulating fine-grained domain factors. Based on these learned prompts, we introduce an adaptive prompt orchestration mechanism that dynamically aggregates these prompts conditioned on current observations. This enables the agent to adaptively construct optimal state representations by identifying dominant domain factors instantaneously. Consequently, the policy optimization is effectively shielded from irrelevant interference, preventing the common issue of overfitting to source domains. Extensive experiments demonstrate that CAPO significantly outperforms state-of-the-art baselines in sample efficiency and asymptotic performance. Crucially, it exhibits superior zero-shot adaptation across unseen target domains characterized by drastic environmental (e.g., illumination) and physical shifts (e.g., field-of-view and rotation), validating its effectiveness as a viable solution for cross-embodiment visuomotor policy adaptation.