Constrained Behavior Cloning for Robotic Learning

TL;DR

This paper introduces a geometrically and historically constrained behavior cloning method that improves robustness and stability in robotic learning by incorporating high-level state information, leading to significant success rate improvements.

Contribution

The paper proposes GHCBC, a novel approach combining geometric and temporal constraints to enhance behavior cloning performance in robotics.

Findings

Success rates increased by 29.73% in simulation.

Success rates increased by 39.4% in real robot experiments.

Significant improvements in long-term operational scenes.

Abstract

Behavior cloning (BC) is a popular supervised imitation learning method in the societies of robotics, autonomous driving, etc., wherein complex skills can be learned by direct imitation from expert demonstrations. Despite its rapid development, it is still affected by limited field of view where accumulation of sensors and joint noise bring compounding errors. In this paper, we introduced geometrically and historically constrained behavior cloning (GHCBC) to dominantly consider high-level state information inspired by neuroscientists, wherein the geometrically constrained behavior cloning were used to geometrically constrain predicting poses, and the historically constrained behavior cloning were utilized to temporally constrain action sequences. The synergy between these two types of constrains enhanced the BC performance in terms of robustness and stability. Comprehensive experimental results showed that success rates were improved by 29.73% in simulation and 39.4% in real robot experiments in average, respectively, compared to state-of-the-art BC method, especially in long-term operational scenes, indicating great potential of using the GHCBC for robotic learning.