Ex-DoF: Expansion of Action Degree-of-Freedom with Virtual Camera Rotation for Omnidirectional Image

TL;DR

This paper introduces a transfer learning method that uses virtual camera rotation to augment data, enabling a robot with fewer DoFs to learn control policies for robots with more DoFs using omnidirectional images.

Contribution

It proposes a novel transfer learning approach leveraging virtual camera rotation for data augmentation in omnidirectional image-based robot control.

Findings

Successful transfer from 3-DoF to 6-DoF robot control policies.

Demonstrated object reaching with a 6-DoF arm using multiple policies.

Effective data augmentation improves learning in high-DoF robotic systems.

Abstract

Inter-robot transfer of training data is a little explored topic in learning- and vision-based robot control. Here we propose a transfer method from a robot with a lower Degree-of-Freedom (DoF) to one with a higher DoF utilizing the omnidirectional camera image. The virtual rotation of the robot camera enables data augmentation in this transfer learning process. As an experimental demonstration, a vision-based control policy for a 6-DoF robot is trained using a dataset collected by a wheeled ground robot with only three DoFs. Towards the application of robotic manipulations, we also demonstrate a control system of a 6-DoF arm robot using multiple policies with different fields of view to enable object reaching tasks.