End-to-end example-based sim-to-real RL policy transfer based on neural stylisation with application to robotic cutting

TL;DR

This paper introduces a novel neural stylisation-based sim-to-real transfer method for reinforcement learning policies, enabling robots to adapt to real-world tasks like cutting with minimal real data and improved stability.

Contribution

It presents a new approach combining neural style transfer and variational autoencoders for effective sim-to-real policy transfer in contact-rich robotic tasks.

Findings

Improved task completion time over baseline methods

Enhanced behavioural stability in real-world deployment

Robustness to geometric and material variations

Abstract

Whereas reinforcement learning has been applied with success to a range of robotic control problems in complex, uncertain environments, reliance on extensive data - typically sourced from simulation environments - limits real-world deployment due to the domain gap between simulated and physical systems, coupled with limited real-world sample availability. We propose a novel method for sim-to-real transfer of reinforcement learning policies, based on a reinterpretation of neural style transfer from image processing to synthesise novel training data from unpaired unlabelled real world datasets. We employ a variational autoencoder to jointly learn self-supervised feature representations for style transfer and generate weakly paired source-target trajectories to improve physical realism of synthesised trajectories. We demonstrate the application of our approach based on the case study of robot cutting of unknown materials. Compared to baseline methods, including our previous work, CycleGAN, and conditional variational autoencoder-based time series translation, our approach achieves improved task completion time and behavioural stability with minimal real-world data. Our framework demonstrates robustness to geometric and material variation, and highlights the feasibility of policy adaptation in challenging contact-rich tasks where real-world reward information is unavailable.