<ul data-eligibleForWebStory="true"><li>A new Time-Unified Diffusion Policy (TUDP) has been developed for robotic manipulation to efficiently generate robot actions with high accuracy.</li><li>The TUDP integrates action recognition capabilities to streamline the action denoising process while enhancing training efficiency and speeding up action generation.</li><li>It introduces a time-unified velocity field in action space with action discrimination information to simplify policy learning and improve action generation speed.</li><li>The TUDP also implements an action-wise training method that includes an action discrimination branch to enhance successful action recognition and denoising accuracy.</li><li>This novel method achieved state-of-the-art performance on RLBench with success rates of 82.6% on a multi-view setup and 83.8% on a single-view setup.</li><li>When using fewer denoising iterations, TUDP demonstrated a significant improvement in success rate, showcasing its efficiency.</li><li>The TUDP is capable of producing accurate actions for various real-world tasks, making it a versatile and reliable solution for robotic manipulation.</li></ul>

Time-Unified Diffusion Policy with Action Discrimination for Robotic Manipulation

Discover more