CDP: Towards Robust Autoregressive Visuomotor Policy Learning via Causal Diffusion

TL;DR

This paper introduces Causal Diffusion Policy (CDP), a transformer-based diffusion model that improves visuomotor policy learning by conditioning on historical actions and employing caching for efficiency, demonstrating robustness in real-world robotic tasks.

Contribution

The paper presents a novel transformer-based diffusion model with a caching mechanism that leverages historical actions for robust and efficient visuomotor policy learning in robotics.

Findings

CDP outperforms existing methods in accuracy across diverse manipulation tasks.

CDP maintains high precision under degraded observation conditions.

The caching mechanism significantly reduces inference computation time.

Abstract

Diffusion Policy (DP) enables robots to learn complex behaviors by imitating expert demonstrations through action diffusion. However, in practical applications, hardware limitations often degrade data quality, while real-time constraints restrict model inference to instantaneous state and scene observations. These limitations seriously reduce the efficacy of learning from expert demonstrations, resulting in failures in object localization, grasp planning, and long-horizon task execution. To address these challenges, we propose Causal Diffusion Policy (CDP), a novel transformer-based diffusion model that enhances action prediction by conditioning on historical action sequences, thereby enabling more coherent and context-aware visuomotor policy learning. To further mitigate the computational cost associated with autoregressive inference, a caching mechanism is also introduced to store attention key-value pairs from previous timesteps, substantially reducing redundant computations during execution. Extensive experiments in both simulated and real-world environments, spanning diverse 2D and 3D manipulation tasks, demonstrate that CDP uniquely leverages historical action sequences to achieve significantly higher accuracy than existing methods. Moreover, even when faced with degraded input observation quality, CDP maintains remarkable precision by reasoning through temporal continuity, which highlights its practical robustness for robotic control under realistic, imperfect conditions.