mlsanityData to 14 Oct 2026

Restoring Noisy Demonstration for Imitation Learning With Diffusion Models

arXiv 2510.14467 ↗First posted 16 Oct 2025cs.ROcited by 1

On the map: Diffusion models → Diffusion and flow policies for robotics

also in 1 more
Abstract (arXiv)

Imitation learning (IL) aims to learn a policy from expert demonstrations and has been applied to various applications. By learning from the expert policy, IL methods do not require environmental interactions or reward signals. However, most existing imitation learning algorithms assume perfect expert demonstrations, but expert demonstrations often contain imperfections caused by errors from human experts or sensor/control system inaccuracies. To address the above problems, this work proposes a filter-and-restore framework to best leverage expert demonstrations with inherent noise. Our proposed method first filters clean samples from the demonstrations and then learns conditional diffusion models to recover the noisy ones. We evaluate our proposed framework and existing methods in various domains, including robot arm manipulation, dexterous manipulation, and locomotion. The experiment results show that our proposed framework consistently outperforms existing methods across all the tasks. Ablation studies further validate the effectiveness of each component and demonstrate the framework's robustness to different noise types and levels. These results confirm the practical applicability of our framework to noisy offline demonstration data.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Diffusion and flow policies for robotics, one of 14 directions in Diffusion models. Builds on 58 of its 105 references that are in the map. First posted on arXiv 16 Oct 2025.

Builds on

58 of 105 references.

Closest ancestors · 5

Same direction, newest first

  1. 2024Diffusion-Reward Adversarial Imitation Learning25 May
  2. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusion7 Mar
  3. 2023Diffusion Model-Augmented Behavioral Cloning26 Feb
  4. 2023Imitating Human Behaviour with Diffusion Models25 Jan
  5. 2022Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning12 Aug

Other references · 53

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  3. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  4. 2020Denoising Diffusion Implicit Modelscited by 9,781
  5. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 9,122
  6. 2018Representation Learning with Contrastive Predictive Codingcited by 7,777
  7. 2021Improved Denoising Diffusion Probabilistic Modelscited by 3,761
  8. 2022Video Diffusion Modelscited by 2,305
  9. 2023Consistency Modelscited by 1,822
  10. 2016OpenAI Gymcited by 1,736
  11. 2016Concrete Problems in AI Safetycited by 1,642
  12. 2016Generative Adversarial Imitation Learningcited by 1,597
38 more
  1. 2021Isaac Gym: High Performance GPU-Based Physics Simulation For Robot Learningcited by 998
  2. 2022Denoising Diffusion Restoration Modelscited by 963
  3. 2018Deep Anomaly Detection with Outlier Exposurecited by 942
  4. 2020DivideMix: Learning with Noisy Labels as Semi-supervised Learningcited by 692
  5. 2018Learning to Reweight Examples for Robust Deep Learningcited by 675
  6. 2019Memorizing Normality to Detect Anomaly: Memory-augmented Deep Autoencoder for Unsupervised Anomaly Detectioncited by 564
  7. 2021CutPaste: Self-Supervised Learning for Anomaly Detection and Localizationcited by 531
  8. 2022Improving Diffusion Models for Inverse Problems using Manifold Constraintscited by 508
  9. 2018Behavioral Cloning from Observationcited by 492
  10. 2017Learning Robust Rewards with Adversarial Inverse Reinforcement Learningcited by 490
  11. 2017End-to-end Driving via Conditional Imitation Learningcited by 485
  12. 2021Implicit Behavioral Cloningcited by 452
  13. 2019Deep Semi-Supervised Anomaly Detectioncited by 291
  14. 2020Learning Memory-guided Normality for Anomaly Detectioncited by 278
  15. 2018Multi-Goal Reinforcement Learning: Challenging Robotics Environments and Request for Researchcited by 260
  16. 2019Extrapolating Beyond Suboptimal Demonstrations via Inverse Reinforcement Learning from Observationscited by 260
  17. 2017Inverse Reward Designcited by 249
  18. 2018Generative Adversarial Imitation from Observationcited by 184
  19. 2023A Survey of Imitation Learning: Algorithms, Recent Developments, and Challengescited by 148
  20. 2020Learning and Evaluating Representations for Deep One-class Classificationcited by 133
  21. 2019Imitation Learning from Imperfect Demonstrationcited by 114
  22. 2022Discriminator-Weighted Offline Imitation Learning from Suboptimal Demonstrationscited by 89
  23. 2022UNICON: Combating Label Noise Through Uniform Selection and Contrastive Learningcited by 77
  24. 2024Diffusion Models, Image Super-Resolution And Everything: A Surveycited by 64
  25. 2023GibbsDDRM: A Partially Collapsed Gibbs Sampler for Solving Blind Inverse Problems with Denoising Diffusion Restorationcited by 59
  26. 2021Imitation Learning: Progress, Taxonomies and Challengescited by 58
  27. 2021Confidence-Aware Imitation Learning from Demonstrations with Varying Optimalitycited by 44
  28. 2019Blind Universal Bayesian Image Denoising with Gaussian Noise Level Learningcited by 31
  29. 2021Improved Deep Reinforcement Learning with Expert Demonstrations for Urban Autonomous Drivingcited by 27
  30. 2020Robust Imitation Learning from Noisy Demonstrationscited by 26
  31. 2019Reinforcement Learning from Imperfect Demonstrations under Soft Expert Guidancecited by 25
  32. 2023Offline Imitation Learning with Suboptimal Demonstrations via Relaxed Distribution Matchingcited by 24
  33. 2020TrustMAE: A Noise-Resilient Defect Classification Framework using Memory-Augmented Auto-Encoders with Trust Regionscited by 21
  34. 2018Imitation Learning with Concurrent Actions in 3D Gamescited by 18
  35. 2022Discriminator-Guided Model-Based Offline Imitation Learningcited by 17
  36. 2021Enhancing Unsupervised Anomaly Detection with Score-Guided Networkcited by 8
  37. 2020On the Guaranteed Almost Equivalence between Imitation Learning from Observation and Demonstrationcited by 6
  38. 2022Diverse Imitation Learning via Self-Organizing Generative Modelscited by 5

50 most cited of 53 All 53 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Language-Critique Imitation Learning from Suboptimal Demonstrationscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2023Good Better Best: Self-Motivated Imitation Learning for noisy Demonstrations
  2. 2023Imitation Learning from Purified Demonstrations
  3. 2024DIDA: Denoised Imitation Learning based on Domain Adaptation
  4. 2025CCDP: Composition of Conditional Diffusion Policies with Guided Sampling
  5. 2023Skill Disentanglement for Imitation Learning from Suboptimal Demonstrations
  6. 2023Selective Sampling and Imitation Learning via Online Regression
  7. 2021Learning from Imperfect Demonstrations from Agents with Varying Dynamics
  8. 2026Policy-Calibrated DAgger: Offline Calibrated Noise Injection for Imitation Learning