DABI: Evaluation of Data Augmentation Methods Using Downsampling in Bilateral Control-Based Imitation Learning with Images

TL;DR

This paper introduces DABI, a data augmentation method using downsampling in bilateral control-based imitation learning with images, significantly increasing data and improving robot manipulation success rates.

Contribution

DABI is a novel data augmentation approach that leverages downsampling of multi-rate sensor data to enhance imitation learning performance in robotics.

Findings

Tenfold increase in data through augmentation

Significant improvement in success rates

Effective stabilization of robot operations

Abstract

Autonomous robot manipulation is a complex and continuously evolving robotics field. This paper focuses on data augmentation methods in imitation learning. Imitation learning consists of three stages: data collection from experts, learning model, and execution. However, collecting expert data requires manual effort and is time-consuming. Additionally, as sensors have different data acquisition intervals, preprocessing such as downsampling to match the lowest frequency is necessary. Downsampling enables data augmentation and also contributes to the stabilization of robot operations. In light of this background, this paper proposes the Data Augmentation Method for Bilateral Control-Based Imitation Learning with Images, called "DABI". DABI collects robot joint angles, velocities, and torques at 1000 Hz, and uses images from gripper and environmental cameras captured at 100 Hz as the basis for data augmentation. This enables a tenfold increase in data. In this paper, we collected just 5 expert demonstration datasets. We trained the bilateral control Bi-ACT model with the unaltered dataset and two augmentation methods for comparative experiments and conducted real-world experiments. The results confirmed a significant improvement in success rates, thereby proving the effectiveness of DABI. For additional material, please check https://mertcookimg.github.io/dabi