Research map / Reinforcement learning / Visual and representation RL
Assessing Transferability From Simulation to Reality for Reinforcement Learning
Fabio Muratore, Michael Gienger, Jan Peters
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add domain randomization (SPOTA or EPOpt) instead of plain PPO on nominal model | PPO · Ball-Balancer sim-to-real | transfer success: failed in all but 2 trials -> most trials transfer | Improved |
| add domain randomization (SPOTA or EPOpt) instead of plain PPO on nominal model | PPO · Cart-Pole sim-to-real | transfer success: failed in all trials -> most trials transfer | Improved |
| add domain randomization instead of nominal-only training | PPO · Ball-Balancer and Cart-Pole sim-to-sim | robustness to parameter changes | Improved |
| use CVaR-based objective (EPOpt) instead of expected return | EPOpt · Sim-to-sim | nominal-domain performance | Hurt |
| add domain randomization instead of nominal training | PPO · Sim-to-sim | variance of returns | Improved |
| increase number of domains during SPOTA iterations (ablation) | SPOTA · Ball-Balancer (simulated) | UCBOG | Improved |
| increase number of domains n in illustrative example (ablation) | Optimal policy for n domains · Illustrative catapult example | simulation optimization bias | Improved |
Closest papers by meaning
- Domain Randomization via Entropy Maximization (ICLR 2024 · similarity 0.89)
- Flow-based Domain Randomization for Learning and Sequencing Robotic Skills (ICML 2025 · similarity 0.89)
- Overcoming the Sim-to-Real Gap: Leveraging Simulation to Learn to Explore for Real-World RL (NeurIPS 2024 · similarity 0.88)
- Rapidly Adapting Policies to the Real-World via Simulation-Guided Fine-Tuning (ICLR 2025 · similarity 0.87)
- SPiDR: A Simple Approach for Zero-Shot Safety in Sim-to-Real Transfer (NeurIPS 2025 · similarity 0.87)
- Generalizable Domain Adaptation for Sim-and-Real Policy Co-Training (NeurIPS 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
