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

arXiv: https://arxiv.org/abs/2601.20846 · HTML: https://mlsanity.com/paper/2601.20846/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in RL for robot locomotion and manipulation, one of 15 directions in Deep reinforcement learning methods.
- Builds on 16 of its 33 references that are in the map.
- First posted on arXiv 28 Jan 2026.

## Placed in

- Deep reinforcement learning methods / [RL for robot locomotion and manipulation](https://mlsanity.com/deeprl/rl-for-robot-locomotion-and-manipulation/)

## Builds on (most cited first, 16 of 16)

- [A Neural Algorithm of Artistic Style](https://mlsanity.com/paper/1508.06576/) (2015)
- [Deep Reconstruction-Classification Networks for Unsupervised Domain Adaptation](https://mlsanity.com/paper/1607.03516/) (2016)
- [Controlling Perceptual Factors in Neural Style Transfer](https://mlsanity.com/paper/1611.07865/) (2016)
- [Variable Impedance Control in End-Effector Space: An Action Space for Reinforcement Learning in Contact-Rich Tasks](https://mlsanity.com/paper/1906.08880/) (2019)
- [Regularizing Action Policies for Smooth Control with Reinforcement Learning](https://mlsanity.com/paper/2012.06644/) (2020)
- [Transfer from Simulation to Real World through Learning Deep Inverse Dynamics Model](https://mlsanity.com/paper/1610.03518/) (2016)
- [Domain Adaptation In Reinforcement Learning Via Latent Unified State Representation](https://mlsanity.com/paper/2102.05714/) (2021)
- [Variable Compliance Control for Robotic Peg-in-Hole Assembly: A Deep Reinforcement Learning Approach](https://mlsanity.com/paper/2008.10224/) (2020)
- [Sim2Real Instance-Level Style Transfer for 6D Pose Estimation](https://mlsanity.com/paper/2203.02069/) (2022)
- [Benchmarking Smoothness and Reducing High-Frequency Oscillations in Continuous Control Policies](https://mlsanity.com/paper/2410.16632/) (2024)
- [Generalized Zero-Shot Domain Adaptation via Coupled Conditional Variational Autoencoders](https://mlsanity.com/paper/2008.01214/) (2020)
- [One-Shot Domain-Adaptive Imitation Learning via Progressive Learning](https://mlsanity.com/paper/2204.11251/) (2022)
- [StyleTime: Style Transfer for Synthetic Time Series Generation](https://mlsanity.com/paper/2209.11306/) (2022)
- [Learning robotic milling strategies based on passive variable operational space interaction control](https://mlsanity.com/paper/2304.01000/) (2023)
- [Neural Policy Style Transfer](https://mlsanity.com/paper/2402.00677/) (2024)
- [Neural Style Transfer with Twin-Delayed DDPG for Shared Control of Robotic Manipulators](https://mlsanity.com/paper/2402.00722/) (2024)

## Abstract (arXiv)

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.