Accept Synthetic Objects as Real: End-to-End Training of Attentive Deep Visuomotor Policies for Manipulation in Clutter

TL;DR

This paper introduces ASOR, a data augmentation method enabling end-to-end training of deep visuomotor policies for robotic manipulation in clutter, significantly improving success rates over previous methods.

Contribution

It presents ASOR, a novel data augmentation technique, and two attention-based network architectures that enhance manipulation performance in cluttered environments.

Findings

ASOR-IA and ASOR-EA outperform previous methods in cluttered settings.

Both architectures perform better than prior approaches in uncluttered environments.

ASOR-EA achieves the best performance overall.

Abstract

Recent research demonstrated that it is feasible to end-to-end train multi-task deep visuomotor policies for robotic manipulation using variations of learning from demonstration (LfD) and reinforcement learning (RL). In this paper, we extend the capabilities of end-to-end LfD architectures to object manipulation in clutter. We start by introducing a data augmentation procedure called Accept Synthetic Objects as Real (ASOR). Using ASOR we develop two network architectures: implicit attention ASOR-IA and explicit attention ASOR-EA. Both architectures use the same training data (demonstrations in uncluttered environments) as previous approaches. Experimental results show that ASOR-IA and ASOR-EA succeed ina significant fraction of trials in cluttered environments where previous approaches never succeed. In addition, we find that both ASOR-IA and ASOR-EA outperform previous approaches even in uncluttered environments, with ASOR-EA performing better even in clutter compared to the previous best baseline in an uncluttered environment.