mlsanityData to 7 Oct 2026

Bottom-Up Skill Discovery from Unsegmented Demonstrations for Long-Horizon Robot Manipulation

arXiv 2109.13841 ↗First posted 28 Sep 2021cs.RO cs.LGcited by 82

On the map: Offline RL and imitation learning from data → Robot manipulation imitation learning

Abstract (arXiv)

We tackle real-world long-horizon robot manipulation tasks through skill discovery. We present a bottom-up approach to learning a library of reusable skills from unsegmented demonstrations and use these skills to synthesize prolonged robot behaviors. Our method starts with constructing a hierarchical task structure from each demonstration through agglomerative clustering. From the task structures of multi-task demonstrations, we identify skills based on the recurring patterns and train goal-conditioned sensorimotor policies with hierarchical imitation learning. Finally, we train a meta controller to compose these skills to solve long-horizon manipulation tasks. The entire model can be trained on a small set of human demonstrations collected within 30 minutes without further annotations, making it amendable to real-world deployment. We systematically evaluated our method in simulation environments and on a real robot. Our method has shown superior performance over state-of-the-art imitation learning methods in multi-stage manipulation tasks. Furthermore, skills discovered from multi-task demonstrations boost the average task success by 8% compared to those discovered from individual tasks.

Key facts. Cited by 82 papers in the map (data through 7 Oct 2026). Placed in Robot manipulation imitation learning, one of 13 directions in Offline RL and imitation learning from data. Builds on 31 of its 59 references that are in the map. First posted on arXiv 28 Sep 2021.

Builds on

31 of 59 references.

Closest ancestors · 5

Same direction, newest first

  1. 2020Learning Multi-Arm Manipulation Through Collaborative Teleoperation12 Dec
  2. 2020Learning to Generalize Across Long-Horizon Tasks from Human Demonstrations13 Mar
  3. 2019Relay Policy Learning: Solving Long-Horizon Tasks via Imitation and Reinforcement Learning25 Oct
  4. 2018CompILE: Compositional Imitation Learning and Execution4 Dec
  5. 2018One-Shot Hierarchical Imitation Learning of Compound Visuomotor Tasks25 Oct

Other references · 26

  1. 2017Hindsight Experience Replaycited by 1,096
  2. 2016Stacked Hourglass Networks for Human Pose Estimationcited by 913
  3. 2018CornerNet: Detecting Objects as Paired Keypointscited by 763
  4. 2018Diversity is All You Need: Learning Skills without a Reward Functioncited by 734
  5. 2020robosuite: A Modular Simulation Framework and Benchmark for Robot Learningcited by 484
  6. 2017FeUdal Networks for Hierarchical Reinforcement Learningcited by 328
  7. 2019Dynamics-Aware Unsupervised Discovery of Skillscited by 308
  8. 2017Deep Imitation Learning for Complex Manipulation Tasks from Virtual Reality Teleoperationcited by 286
  9. 2020Integrated Task and Motion Planningcited by 281
  10. 2016Variational Intrinsic Controlcited by 262
  11. 2017One-Shot Imitation Learningcited by 232
  12. 2019Bottom-up Object Detection by Grouping Extreme and Center Pointscited by 156
14 more
  1. 2019AVID: Learning Multi-Stage Tasks via Pixel-Level Translation of Human Videoscited by 105
  2. 2019Efficient Parameter-free Clustering Using First Neighbor Relationscited by 96
  3. 2018Hierarchical Imitation and Reinforcement Learningcited by 88
  4. 2019IRIS: Implicit Reinforcement without Interaction at Scale for Learning Control from Offline Robot Manipulation Datacited by 85
  5. 2019Making Sense of Vision and Touch: Learning Multimodal Representations for Contact-Rich Taskscited by 80
  6. 2017Neural Task Programming: Learning to Generalize Across Hierarchical Taskscited by 75
  7. 2018TACO: Learning Task Decomposition via Temporal Alignment for Controlcited by 61
  8. 2021Temporally-Weighted Hierarchical Clustering for Unsupervised Action Segmentationcited by 58
  9. 2020Learning Robot Skills with Temporal Variational Inferencecited by 54
  10. 2017Multi-Level Discovery of Deep Optionscited by 50
  11. 2021SKID RAW: Skill Discovery from Raw Trajectoriescited by 22
  12. 2020Modeling Long-horizon Tasks as Sequential Interaction Landscapescited by 14
  13. 2015Action Recognition by Hierarchical Mid-level Action Elementscited by 6
  14. 2015A Bottom-up Approach for Pancreas Segmentation using Cascaded Superpixels and (Deep) Image Patch Labelingcited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 7; 2023: 6; 2024: 19; 2025: 28; 2026: 22.

82 papers.

Sort:

By area:

  1. 2024OpenVLA: An Open-Source Vision-Language-Action Modelcited by 2,777
  2. 2024Octo: An Open-Source Generalist Robot Policycited by 1,427
  3. 2024RDT-1B: a Diffusion Foundation Model for Bimanual Manipulationcited by 670
  4. 2025SpatialVLA: Exploring Spatial Representations for Visual-Language-Action Modelcited by 502
  5. 2024CogACT: A Foundational Vision-Language-Action Model for Synergizing Cognition and Action in Robotic Manipulationcited by 411
  6. 2025UniVLA: Learning to Act Anywhere with Task-centric Latent Actionscited by 386
  7. 2023MimicGen: A Data Generation System for Scalable Robot Learning using Human Demonstrationscited by 308
  8. 2024RT-H: Action Hierarchies Using Languagecited by 204
  9. 2025HybridVLA: Collaborative Diffusion and Autoregression in a Unified Vision-Language-Action Modelcited by 176
  10. 2024iVideoGPT: Interactive VideoGPTs are Scalable World Modelscited by 136
  11. 2022VIOLA: Imitation Learning for Vision-Based Manipulation with Object Proposal Priorscited by 135
  12. 2024Scaling Cross-Embodied Learning: One Policy for Manipulation, Navigation, Locomotion and Aviationcited by 108
38 more
  1. 2024DexMimicGen: Automated Data Generation for Bimanual Dexterous Manipulation via Imitation Learningcited by 104
  2. 2022Learning and Retrieval from Prior Data for Skill-based Imitation Learningcited by 77
  3. 2023XSkill: Cross Embodiment Skill Discoverycited by 74
  4. 2024IGOR: Image-GOal Representations are the Atomic Control Units for Foundation Models in Embodied AIcited by 68
  5. 2023LOTUS: Continual Imitation Learning for Robot Manipulation Through Unsupervised Skill Discoverycited by 63
  6. 2022Learning Neuro-Symbolic Skills for Bilevel Planningcited by 55
  7. 2026Robometer: Scaling General-Purpose Robotic Reward Models via Trajectory Comparisonscited by 51
  8. 2024SkillMimicGen: Automated Demonstration Generation for Efficient Skill Learning and Deploymentcited by 48
  9. 2025Sim-and-Real Co-Training: A Simple Recipe for Vision-Based Robotic Manipulationcited by 42
  10. 2024A Survey of Embodied Learning for Object-Centric Robotic Manipulationcited by 34
  11. 2022LEAGUE: Guided Skill Learning and Abstraction for Long-Horizon Manipulationcited by 32
  12. 2022Eliciting Compatible Demonstrations for Multi-Human Imitation Learningcited by 22
  13. 2022Learning Options via Compressioncited by 22
  14. 2025MLA: A Multisensory Language-Action Model for Multimodal Understanding and Forecasting in Robotic Manipulationcited by 22
  15. 2026LAP: Language-Action Pre-Training Enables Zero-shot Cross-Embodiment Transfercited by 17
  16. 2025Learning Compositional Behaviors from Demonstration and Languagecited by 15
  17. 2025XR-1: Towards Versatile Vision-Language-Action Models via Learning Unified Vision-Motion Representationscited by 15
  18. 2026CLARE: Continual Learning for Vision-Language-Action Models via Autonomous Adapter Routing and Expansioncited by 14
  19. 2025Exploring Embodied Multimodal Large Models: Development, Datasets, and Future Directionscited by 12
  20. 2025How to Train Your Robots? The Impact of Demonstration Modality on Imitation Learningcited by 11
  21. 2026Towards Long-Lived Robots: Continual Learning VLA Models via Reinforcement Fine-Tuningcited by 11
  22. 2024M2Distill: Multi-Modal Distillation for Lifelong Imitation Learningcited by 10
  23. 2025MTIL: Encoding Full History with Mamba for Temporal Imitation Learningcited by 10
  24. 2025TwinVLA: Data-Efficient Bimanual Manipulation with Twin Single-Arm Vision-Language-Action Modelscited by 10
  25. 2025HiMoE-VLA: Hierarchical Mixture-of-Experts for Generalist Vision-Language-Action Policiescited by 10
  26. 2025OXE-AugE: A Large-Scale Robot Augmentation of OXE for Scaling Cross-Embodiment Policy Learningcited by 10
  27. 2026MMaDA-VLA: Large Diffusion Vision-Language-Action Model with Unified Multi-Modal Instruction and Generationcited by 10
  28. 2024Disentangled Unsupervised Skill Discovery for Efficient Hierarchical Reinforcement Learningcited by 9
  29. 2025SPECI: Skill Prompts based Hierarchical Continual Imitation Learning for Robot Manipulationcited by 9
  30. 2025MoE-DP: An MoE-Enhanced Diffusion Policy for Robust Long-Horizon Robotic Manipulation with Skill Decomposition and Failure Recoverycited by 9
  31. 2024PRIME: Scaffolding Manipulation Tasks with Behavior Primitives for Data-Efficient Imitation Learningcited by 8
  32. 2025BOSS: Benchmark for Observation Space Shift in Long-Horizon Taskcited by 7
  33. 2025Open-World Skill Discovery from Unsegmented Demonstrationscited by 7
  34. 2025LatBot: Distilling Universal Latent Actions for Vision-Language-Action Modelscited by 7
  35. 2026GuidedVLA: Specifying Task-Relevant Factors via Plug-and-Play Action Attention Specializationcited by 7
  36. 2026RotVLA: Rotational Latent Action for Vision-Language-Action Modelcited by 7
  37. 2026Hierarchical Latent Action Modelcited by 6
  38. 2024BiKC: Keypose-Conditioned Consistency Policy for Bimanual Robotic Manipulationcited by 5

50 most cited of 82 All 82 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026Offline Discovery of Interpretable Skills from Multi-Task Trajectories
  2. 2026SADP: Subgoal-Aware Diffusion Policy for Long-Horizon Manipulation Learned from Foundation Model Generated Demonstrations
  3. 2023MimicPlay: Long-Horizon Imitation Learning by Watching Human Play
  4. 2026Beyond the Majority: Long-tail Imitation Learning for Robotic Manipulation
  5. 2025ManiLong-Shot: Interaction-Aware One-Shot Imitation Learning for Long-Horizon Manipulation
  6. 2025Multi-Stage Manipulation with Demonstration-Augmented Reward, Policy, and World Model Learning
  7. 2025Bootstrapping Imitation Learning for Long-horizon Manipulation via Hierarchical Data Collection Space
  8. 2022Latent Plans for Task-Agnostic Offline Reinforcement Learning