mlsanityData to 7 Oct 2026

Diff-DAgger: Uncertainty Estimation with Diffusion Policy for Robotic Manipulation

arXiv 2410.14868 ↗First posted 18 Oct 2024cs.ROcited by 29

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

Abstract (arXiv)

Recently, diffusion policy has shown impressive results in handling multi-modal tasks in robotic manipulation. However, it has fundamental limitations in out-of-distribution failures that persist due to compounding errors and its limited capability to extrapolate. One way to address these limitations is robot-gated DAgger, an interactive imitation learning with a robot query system to actively seek expert help during policy rollout. While robot-gated DAgger has high potential for learning at scale, existing methods like Ensemble-DAgger struggle with highly expressive policies: They often misinterpret policy disagreements as uncertainty at multi-modal decision points. To address this problem, we introduce Diff-DAgger, an efficient robot-gated DAgger algorithm that leverages the training objective of diffusion policy. We evaluate Diff-DAgger across different robot tasks including stacking, pushing, and plugging, and show that Diff-DAgger improves the task failure prediction by 39.0%, the task completion rate by 20.6%, and reduces the wall-clock time by a factor of 7.8. We hope that this work opens up a path for efficiently incorporating expressive yet data-hungry policies into interactive robot learning settings. The project website is available at: https://diffdagger.github.io.

Key facts. Cited by 29 papers in the map (data through 7 Oct 2026). Placed in Diffusion and flow policies for robotics, one of 14 directions in Diffusion models. Builds on 25 of its 32 references that are in the map. First posted on arXiv 18 Oct 2024.

Builds on

25 of 32 references.

Closest ancestors · 5

Same direction, newest first

  1. 2023DiffAIL: Diffusion Adversarial Imitation Learning11 Dec
  2. 2023Adaptive Online Replanning with Diffusion Models14 Oct
  3. 2023Motion Planning Diffusion: Learning and Planning of Robot Motions with Diffusion Models3 Aug
  4. 2023Diffusion Policies for Out-of-Distribution Generalization in Offline Reinforcement Learning10 Jul
  5. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusion7 Mar

Other references · 20

  1. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  2. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  3. 2017Proximal Policy Optimization Algorithmscited by 14,441
  4. 2017FiLM: Visual Reasoning with a General Conditioning Layercited by 2,286
  5. 2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardwarecited by 1,490
  6. 2018On the Continuity of Rotation Representations in Neural Networkscited by 1,186
  7. 2015End-to-End Training of Deep Visuomotor Policiescited by 847
  8. 2021What Matters in Learning from Offline Human Demonstrations for Robot Manipulationcited by 699
  9. 2022R3M: A Universal Visual Representation for Robot Manipulationcited by 678
  10. 2024Universal Manipulation Interface: In-The-Wild Robot Teaching Without In-The-Wild Robotscited by 373
  11. 2023Your Diffusion Model is Secretly a Zero-Shot Classifiercited by 301
  12. 2018HG-DAgger: Interactive Imitation Learning with Human Expertscited by 245
8 more
  1. 2024ManiSkill3: GPU Parallelized Robotics Simulation and Rendering for Generalizable Embodied AIcited by 216
  2. 2021ManiSkill: Generalizable Manipulation Skill Benchmark with Large-Scale Demonstrationscited by 193
  3. 2021ThriftyDAgger: Budget-Aware Novelty and Risk Gating for Interactive Imitation Learningcited by 98
  4. 2018EnsembleDAgger: A Bayesian Approach to Safe Imitation Learningcited by 96
  5. 2021Error-Aware Imitation Learning from Teleoperation Data for Mobile Manipulationcited by 59
  6. 2023Model-Based Runtime Monitoring with Interactive Imitation Learningcited by 37
  7. 2022Fleet-DAgger: Interactive Robot Fleet Learning with Scalable Human Supervisioncited by 29
  8. 2022PATO: Policy Assisted TeleOperation for Scalable Robot Data Collectioncited by 19

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 11; 2026: 18.

29 papers.

Sort:

By area:

  1. 2025Failure Prediction at Runtime for Generative Robot Policiescited by 36
  2. 2026When to Trust Imagination: Adaptive Action Execution for World Action Modelscited by 18
  3. 2025Demystifying Diffusion Policies: Action Memorization and Simple Lookup Table Alternativescited by 9
  4. 2025Dynamic Rank Adjustment in Diffusion Policies for Efficient and Flexible Trainingcited by 6
  5. 2025Uncertainty Comes for Free: Human-in-the-Loop Policies with Diffusion Modelscited by 6
  6. 2025Accelerating Residual Reinforcement Learning with Uncertainty Estimationcited by 6
  7. 2026FlowCorrect: Efficient Interactive Correction of Generative Flow Policies for Robotic Manipulationcited by 5
  8. 2025DAgger Diffusion Navigation: DAgger Boosted Diffusion Policy for Vision-Language Navigationcited by 4
  9. 2026Uncertainty Quantification for Flow-Based Vision-Language-Action Modelscited by 4
  10. 2025Interactive Imitation Learning for Dexterous Robotic Manipulation: Challenges and Perspectives -- A Surveycited by 3
  11. 2026Generative Control as Optimization: Time Unconditional Flow Matching for Adaptive and Robust Robotic Controlcited by 3
  12. 2025From Action Labels to Sets: Rethinking Action Supervision for Imitation Learning from Corrective Feedbackcited by 2
17 more
  1. 2025PANDORA: Diffusion Policy Learning for Dexterous Robotic Piano Playingcited by 2
  2. 2026Dreaming the Unseen: World Model-regularized Diffusion Policy for Out-of-Distribution Robustnesscited by 2
  3. 2026Hand-in-the-Loop: Improving VLA Policies for Dexterous Manipulation via Seamless Hand-Arm Interventioncited by 2
  4. 2026ActProbe: Action-Space Probe for Early Failure Detection of Generative Robot Policiescited by 2
  5. 2026VLA-FAIL: Efficient Task Failure Detection for Finetuned Vision-Language-Action Modelscited by 2
  6. 2026RECALL: Recovery Experience Collection for Active Lifelong Learning in Vision-Language-Action Modelscited by 2
  7. 2026The Geometry of Flow-Matching Uncertainty: A Cost-free Uncertainty Proxy and Its Application in Flow-based VLA Failure Detectioncited by 2
  8. 2025Improving Robustness to Out-of-Distribution States in Imitation Learning via Deep Koopman-Boosted Diffusion Policycited by 1
  9. 2026Set-Supervised Diffusion Policy: Learning Action-Chunking Diffusion through Correctionscited by 1
  10. 2026DISEIL: Demonstration Distillation for Sample-Efficient Imitation Learningcited by 1
  11. 2025ASkDAgger: Active Skill-level Data Aggregation for Interactive Imitation Learningcited by 0
  12. 2026VoxelDiffusionCut: Non-destructive Internal-part Extraction via Iterative Cutting and Structure Estimationcited by 0
  13. 2026Act, Think or Abstain: Complexity-Aware Adaptive Inference for Vision-Language-Action Modelscited by 0
  14. 2026Easy-IIL: Reducing Human Operational Burden in Interactive Imitation Learning via Assistant Expertscited by 0
  15. 2026ReflectVLN: Training Vision-Language Navigation Agents with Reflective Reasoningcited by 0
  16. 2026Making two action heads agree: coordination mechanisms and a runtime collapse certificate for flow-matching policiescited by 0
  17. 2026Policy-Calibrated DAgger: Offline Calibrated Noise Injection for Imitation Learningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Enabling Stateful Behaviors for Diffusion-based Policy Learning
  2. 2025D2PPO: Diffusion Policy Policy Optimization with Dispersive Loss
  3. 2024A context-adaptive policy framework for robust and reactive robotic manipulation via uncertainty-aware imitation learning
  4. 2026Abstracting Robot Manipulation Skills via Mixture-of-Experts Diffusion Policies
  5. 2025Flexible Multitask Learning with Factorized Diffusion Policy
  6. 2025Two-Steps Diffusion Policy for Robotic Manipulation via Genetic Denoising
  7. 2025Hybrid Consistency Policy: Decoupling Multi-Modal Diversity and Real-Time Efficiency in Robotic Manipulation
  8. 2025CCDP: Composition of Conditional Diffusion Policies with Guided Sampling