English

Bridging the Sim2Real Gap: Vision Encoder Pre-Training for Visuomotor Policy Transfer

Robotics 2025-09-09 v2 Computer Vision and Pattern Recognition

Abstract

Simulation offers a scalable and efficient alternative to real-world data collection for learning visuomotor robotic policies. However, the simulation-to-reality, or Sim2Real distribution shift -- introduced by employing simulation-trained policies in real-world environments -- frequently prevents successful policy transfer. We present an offline framework to evaluate the performance of using large-scale pre-trained vision encoders to address the Sim2Real gap. We examine a diverse collection of encoders, assessing their ability to extract features necessary for robot control (Action Score) while remaining invariant to task-irrelevant environmental variations (Domain Invariance Score). Evaluating 23 encoders, we reveal patterns across architectures, pre-training datasets, and parameter scales. Our findings show that manipulation-pretrained encoders consistently achieve higher Action Scores, CNN-based encoders demonstrate stronger domain invariance than ViTs, and the best-performing models combine both properties, underscoring DIS and AS as complementary predictors of Sim2Real transferability.

Keywords

Cite

@article{arxiv.2501.16389,
  title  = {Bridging the Sim2Real Gap: Vision Encoder Pre-Training for Visuomotor Policy Transfer},
  author = {Yash Yardi and Samuel Biruduganti and Lars Ankile},
  journal= {arXiv preprint arXiv:2501.16389},
  year   = {2025}
}

Comments

6 pages, 4 figures, 1 table, GitHub: https://github.com/yyardi/Bridging-the-Sim2Real-Gap

R2 v1 2026-06-28T21:20:29.446Z