mlsanityData to 7 Oct 2026

DemoStart: Demonstration-led auto-curriculum applied to sim-to-real with multi-fingered robots

arXiv 2409.06613 ↗First posted 10 Sep 2024cs.RO cs.LGcited by 7

Not on the map. Why?

Abstract (arXiv)

We present DemoStart, a novel auto-curriculum reinforcement learning method capable of learning complex manipulation behaviors on an arm equipped with a three-fingered robotic hand, from only a sparse reward and a handful of demonstrations in simulation. Learning from simulation drastically reduces the development cycle of behavior generation, and domain randomization techniques are leveraged to achieve successful zero-shot sim-to-real transfer. Transferred policies are learned directly from raw pixels from multiple cameras and robot proprioception. Our approach outperforms policies learned from demonstrations on the real robot and requires 100 times fewer demonstrations, collected in simulation. More details and videos in https://sites.google.com/view/demostart.

Key facts. Cited by 7 papers in the map (data through 7 Oct 2026). Builds on 39 of its 55 references that are in the map. First posted on arXiv 10 Sep 2024.

Builds on

39 of 55 references.

References · 39

  1. 2016Layer Normalizationcited by 5,579
  2. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusioncited by 2,730
  3. 2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardwarecited by 1,490
  4. 2017Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real Worldcited by 1,401
  5. 2018IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architecturescited by 866
  6. 2019Solving Rubik's Cube with a Robot Handcited by 699
  7. 2017Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrationscited by 685
  8. 2018Learning Dexterous In-Hand Manipulationcited by 643
  9. 2019Learning by Cheatingcited by 361
  10. 2017Overcoming Exploration in Reinforcement Learning with Demonstrationscited by 322
  11. 2018Maximum a Posteriori Policy Optimisationcited by 318
  12. 2017Leveraging Demonstrations for Deep Reinforcement Learning on Robotics Problems with Sparse Rewardscited by 302
27 more
  1. 2020Emergent Complexity and Zero-shot Transfer via Unsupervised Environment Designcited by 260
  2. 2020Prioritized Level Replaycited by 190
  3. 2017Reverse Curriculum Generation for Reinforcement Learningcited by 185
  4. 2021Open-Ended Learning Leads to Generally Capable Agentscited by 177
  5. 2018Learning by Playing - Solving Sparse Reward Tasks from Scratchcited by 164
  6. 2020Automatic Curriculum Learning For Deep RL: A Short Surveycited by 141
  7. 2022DeXtreme: Transfer of Agile In-hand Manipulation from Simulation to Realitycited by 136
  8. 2023Human-Timescale Adaptation in an Open-Ended Task Spacecited by 136
  9. 2023RoboCat: A Self-Improving Generalist Agent for Robotic Manipulationcited by 108
  10. 2023UniDexGrasp++: Improving Dexterous Grasping Policy Learning via Geometry-aware Curriculum and Iterative Generalist-Specialist Learningcited by 101
  11. 2023General In-Hand Object Rotation with Vision and Touchcited by 87
  12. 2023Rotating without Seeing: Towards In-hand Dexterity through Touchcited by 78
  13. 2023LEAP Hand: Low-Cost, Efficient, and Anthropomorphic Hand for Robot Learningcited by 78
  14. 2018Learning Montezuma's Revenge from a Single Demonstrationcited by 76
  15. 2017Data-efficient Deep Reinforcement Learning for Dexterous Manipulationcited by 61
  16. 2024ALOHA 2: An Enhanced Low-Cost Hardware for Bimanual Teleoperationcited by 61
  17. 2023IndustReal: Transferring Contact-Rich Assembly Tasks from Simulation to Realitycited by 44
  18. 2021Transferring Dexterous Manipulation from GPU Simulation to a Remote Real-World TriFingercited by 34
  19. 2024Offline Actor-Critic Reinforcement Learning Scales to Large Modelscited by 30
  20. 2023DexPBT: Scaling up Dexterous Manipulation for Hand-Arm Systems with Population Based Trainingcited by 24
  21. 2022Complex In-Hand Manipulation via Compliance-Enabled Finger Gaiting and Multi-Modal Planningcited by 21
  22. 2024Embodied AI with Two Arms: Zero-shot Learning, Safety and Modularitycited by 19
  23. 2018Backplay: "Man muss immer umkehren"cited by 16
  24. 2023DEFT: Dexterous Fine-Tuning for Real-World Hand Policiescited by 15
  25. 2024Reverse Forward Curriculum Learning for Extreme Sample and Demonstration Efficiency in Reinforcement Learningcited by 12
  26. 2024Learning a Shape-Conditioned Agent for Purely Tactile In-Hand Manipulation of Various Objectscited by 7
  27. 2023Dexterous In-hand Manipulation by Guiding Exploration with Simple Sub-skill Controllerscited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 3; 2026: 4.

7 papers.

  1. 2025Sim-to-Real Reinforcement Learning for Vision-Based Dexterous Manipulation on Humanoidscited by 40
  2. 2025Exploiting Policy Idling for Dexterous Manipulationcited by 1
  3. 2026Play2Perfect: What Matters in Dexterous Play Pretraining for Precise Assembly?cited by 1
  4. 2026Robust In-Hand Manipulation via Priors in Reinforcement Learning and Mechanical Designcited by 1
  5. 2025Learning Pivoting Manipulation with Force and Vision Feedback Using Optimization-based Demonstrationscited by 0
  6. 2026ADEPT: Accelerating Dexterity via Pre-Training and Post-Training using Reinforcement Learningcited by 0
  7. 2026X-Reset: Scaling Object-Centric Reinforcement Learning via Cross-Embodiment Resetscited by 0