mlsanityData to 7 Oct 2026

Compliant Residual DAgger: Improving Real-World Contact-Rich Manipulation with Human Corrections

arXiv 2506.16685 ↗First posted 20 Jun 2025cs.RO cs.LGcited by 25

Not on the map. Why?

Abstract (arXiv)

We address key challenges in Dataset Aggregation (DAgger) for real-world contact-rich manipulation: how to collect informative human correction data and how to effectively update policies with this new data. We introduce Compliant Residual DAgger (CR-DAgger), which contains two novel components: 1) a Compliant Intervention Interface that leverages compliance control, allowing humans to provide gentle, accurate delta action corrections without interrupting the ongoing robot policy execution; and 2) a Compliant Residual Policy formulation that learns from human corrections while incorporating force feedback and force control. Our system significantly enhances performance on precise contact-rich manipulation tasks using minimal correction data, improving base policy success rates by 64% on four challenging tasks (book flipping, belt assembly, cable routing, and gear insertion) while outperforming both retraining-from-scratch and finetuning approaches. Through extensive real-world experiments, we provide practical guidance for implementing effective DAgger in real-world robot learning tasks. Result videos are available at: https://compliant-residual-dagger.github.io

Key facts. Cited by 25 papers in the map (data through 7 Oct 2026). Builds on 48 of its 66 references that are in the map. First posted on arXiv 20 Jun 2025.

Builds on

48 of 66 references.

References · 48

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  2. 2022Flow Matching for Generative Modelingcited by 5,426
  3. 2016Overcoming catastrophic forgetting in neural networkscited by 4,718
  4. 2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actorcited by 4,635
  5. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,777
  6. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusioncited by 2,730
  7. 2024π0: A Vision-Language-Action Flow Model for General Robot Controlcited by 2,297
  8. 2016WaveNet: A Generative Model for Raw Audiocited by 1,710
  9. 2024Octo: An Open-Source Generalist Robot Policycited by 1,427
  10. 2023Open X-Embodiment: Robotic Learning Datasets and RT-X Modelscited by 849
  11. 2024DROID: A Large-Scale In-The-Wild Robot Manipulation Datasetcited by 801
  12. 2013An Empirical Investigation of Catastrophic Forgetting in Gradient-Based Neural Networkscited by 711
36 more
  1. 2023DPOK: Reinforcement Learning for Fine-tuning Text-to-Image Diffusion Modelscited by 431
  2. 2024Universal Manipulation Interface: In-The-Wild Robot Teaching Without In-The-Wild Robotscited by 373
  3. 2016Benchmarking Deep Reinforcement Learning for Continuous Controlcited by 361
  4. 2023Efficient Online Reinforcement Learning with Offline Datacited by 332
  5. 2021Deep ViT Features as Dense Visual Descriptorscited by 327
  6. 2018HG-DAgger: Interactive Imitation Learning with Human Expertscited by 245
  7. 2018Residual Reinforcement Learning for Robot Controlcited by 242
  8. 2024Precise and Dexterous Robotic Manipulation via Human-in-the-Loop Reinforcement Learningcited by 153
  9. 2025ConRFT: A Reinforced Fine-tuning Method for VLA Models via Consistency Policycited by 139
  10. 2024Flow as the Cross-Domain Manipulation Interfacecited by 126
  11. 2024Track2Act: Predicting Point Tracks from Internet Videos enables Generalizable Robot Manipulationcited by 116
  12. 2020Human-in-the-Loop Imitation Learning using Remote Teleoperationcited by 103
  13. 2022Robot Learning on the Job: Human-in-the-Loop Autonomy and Learning During Deploymentcited by 101
  14. 2021ThriftyDAgger: Budget-Aware Novelty and Risk Gating for Interactive Imitation Learningcited by 98
  15. 2024From Imitation to Refinement -- Residual RL for Precise Assemblycited by 80
  16. 2023Teach a Robot to FISH: Versatile Imitation from One Minute of Demonstrationscited by 74
  17. 2023XSkill: Cross Embodiment Skill Discoverycited by 74
  18. 2024Policy Agnostic RL: Offline RL and Online RL Fine-Tuning of Any Class and Backbonecited by 71
  19. 2022Watch and Match: Supercharging Imitation with Regularized Optimal Transportcited by 64
  20. 2024Policy Decorator: Model-Agnostic Online Refinement for Large Policy Modelcited by 60
  21. 2021LazyDAgger: Reducing Context Switching in Interactive Imitation Learningcited by 50
  22. 2024Adaptive Compliance Policy: Learning Approximate Compliance for Diffusion Guided Controlcited by 36
  23. 2023See to Touch: Learning Tactile Dexterity through Visual Incentivescited by 33
  24. 2025Vision in Action: Learning Active Perception from Human Demonstrationscited by 33
  25. 2024Bridging the Human to Robot Dexterity Gap through Object-Oriented Rewardscited by 25
  26. 2025RoboCopilot: Human-in-the-loop Interactive Imitation Learning for Robot Manipulationcited by 22
  27. 2025Rapidly Adapting Policies to the Real World via Simulation-Guided Fine-Tuningcited by 15
  28. 2023JacobiNeRF: NeRF Shaping with Mutual Information Gradientscited by 10
  29. 2025RoboPanoptes: The All-seeing Robot with Whole-body Dexteritycited by 9
  30. 2022ASPiRe:Adaptive Skill Priors for Reinforcement Learningcited by 7
  31. 2019Robust Execution of Contact-Rich Motion Plans by Hybrid Force-Velocity Controlcited by 6
  32. 2025Uncertainty Comes for Free: Human-in-the-Loop Policies with Diffusion Modelscited by 6
  33. 2020Manipulation with Shared Graspingcited by 5
  34. 2022Enhancing Generalizable 6D Pose Tracking of an In-Hand Object with Tactile Sensingcited by 5
  35. 2023simPLE: a visuotactile method learned in simulation to precisely pick, localize, regrasp, and place objectscited by 5
  36. 2025CoinFT: A Coin-Sized, Capacitive 6-Axis Force Torque Sensor for Robotic Applicationscited by 5

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 3; 2026: 22.

25 papers.

Sort:

By area:

  1. 2025RaC: Robot Learning for Long-Horizon Tasks by Scaling Recovery and Correctioncited by 35
  2. 2026Uncertainty Quantification in LLM Agents: Foundations, Emerging Challenges, and Opportunitiescited by 15
  3. 2025BAgger: Backwards Aggregation for Mitigating Drift in Autoregressive Video Diffusion Modelscited by 13
  4. 2025ImplicitRDP: An End-to-End Visual-Force Diffusion Policy with Structural Slow-Fast Learningcited by 8
  5. 2026FlowCorrect: Efficient Interactive Correction of Generative Flow Policies for Robotic Manipulationcited by 5
  6. 2026UniIntervene: Agentic Intervention for Efficient Real-World Reinforcement Learningcited by 4
  7. 2026FAR: Failure-Aware Retry for Test-Time Recovery and Continual Policy Improvementcited by 4
  8. 2026Data Pyramid for Embodied Manipulation: A Surveycited by 4
  9. 2026What Makes Value Learning Efficient in Residual Reinforcement Learning?cited by 3
  10. 2026FlowDAgger: Human-in-the-Loop Adaptation of Generative Robot Policies in Latent Spacecited by 3
  11. 2026Hand-in-the-Loop: Improving VLA Policies for Dexterous Manipulation via Seamless Hand-Arm Interventioncited by 2
  12. 2026FA-RDP: A Frequency-Adaptive Reactive Diffusion Policy for Contact-Rich Manipulationcited by 2
13 more
  1. 2026Set-Supervised Diffusion Policy: Learning Action-Chunking Diffusion through Correctionscited by 1
  2. 2026Universal Manipulation Exoskeleton: Learning Compliant Whole-body Policies with Real-time Torque Feedbackcited by 1
  3. 2026Visual Verification Enables Inference-time Steering and Autonomous Policy Improvementcited by 1
  4. 2026OmniTacTune: Policy-Agnostic Real-World RL for Tactile Residual Adaptation of Visual Policiescited by 1
  5. 2026IMPACT: Learning Internal-Model Predictive Control for Forceful Robotic Manipulationcited by 0
  6. 2026HiL-ResRL: A Model-Agnostic Finetuning Adapter via Human-in-the-loop Residual Reinforcement Learningcited by 0
  7. 2026ReGuide: From Test-Time Guidance to Self-Improving Diffusion Policiescited by 0
  8. 2026Robustness of Robotic Manipulation: Foundations and Frontierscited by 0
  9. 2026Asynchronous Multimodal Diffusion Policy Composition via Latency-Aware Guidance Fusioncited by 0
  10. 2026VIDP: Variable Impedance Diffusion Policy for Compliant Robot Manipulation from Diverse Demonstrationscited by 0
  11. 2026ForceDelta-VLA: Distilling Force-Conditioned ActionCorrections for Contact-Rich Manipulationcited by 0
  12. 2026Policy-Calibrated DAgger: Offline Calibrated Noise Injection for Imitation Learningcited by 0
  13. 2026BlenDAgger: Blended Shared Control for Interactive Imitation Learningcited by 0