Paper ID: 2310.20414

Meta Learning for Multi-View Visuomotor Systems

Benji Alwis

This paper introduces a new approach for quickly adapting a multi-view visuomotor system for robots to varying camera configurations from the baseline setup. It utilises meta-learning to fine-tune the perceptual network while keeping the policy network fixed. Experimental results demonstrate a significant reduction in the number of new training episodes needed to attain baseline performance.

Submitted: Oct 31, 2023