mlsanityData to 14 Oct 2026

UniSkill: Imitating Human Videos via Cross-Embodiment Skill Representations

arXiv 2505.08787 ↗First posted 13 May 2025cs.RO cs.CVcited by 35

Not on the map. Why?

Abstract (arXiv)

Mimicry is a fundamental learning mechanism in humans, enabling individuals to learn new tasks by observing and imitating experts. However, applying this ability to robots presents significant challenges due to the inherent differences between human and robot embodiments in both their visual appearance and physical capabilities. While previous methods bridge this gap using cross-embodiment datasets with shared scenes and tasks, collecting such aligned data between humans and robots at scale is not trivial. In this paper, we propose UniSkill, a novel framework that learns embodiment-agnostic skill representations from large-scale cross-embodiment video data without any labels, enabling skills extracted from human video prompts to effectively transfer to robot policies trained only on robot data. Our experiments in both simulation and real-world environments show that our cross-embodiment skills successfully guide robots in selecting appropriate actions, even with unseen video prompts. The project website can be found at: https://kimhanjung.github.io/UniSkill.

Key facts. Cited by 35 papers in the map (data through 14 Oct 2026). Builds on 31 of its 31 references that are in the map. First posted on arXiv 13 May 2025.

Builds on

31 of 31 references.

References · 31

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2020Denoising Diffusion Implicit Modelscited by 9,781
  3. 2020Unsupervised Learning of Visual Features by Contrasting Cluster Assignmentscited by 2,857
  4. 2023Diffusion Policy: Visuomotor Policy Learning via Action Diffusioncited by 2,838
  5. 2022InstructPix2Pix: Learning to Follow Image Editing Instructionscited by 2,611
  6. 2013Sinkhorn Distances: Lightspeed Computation of Optimal Transportation Distancescited by 2,316
  7. 2023Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardwarecited by 1,568
  8. 2024Depth Anything V2cited by 1,422
  9. 2023LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learningcited by 1,410
  10. 2017Hindsight Experience Replaycited by 1,136
  11. 2017The "something something" video database for learning and evaluating visual common sensecited by 1,103
  12. 2024DROID: A Large-Scale In-The-Wild Robot Manipulation Datasetcited by 852
19 more
  1. 2024Genie: Generative Interactive Environmentscited by 747
  2. 2021What Matters in Learning from Offline Human Demonstrations for Robot Manipulationcited by 725
  3. 2022BC-Z: Zero-Shot Task Generalization with Robotic Imitation Learningcited by 523
  4. 2024Mobile ALOHA: Learning Bimanual Mobile Manipulation with Low-Cost Whole-Body Teleoperationcited by 391
  5. 2024Latent Action Pretraining from Videoscited by 321
  6. 2021H2O: Two Hands Manipulating Objects for First Person Interaction Recognitioncited by 254
  7. 2021Bridge Data: Boosting Generalization of Robotic Skills with Cross-Domain Datasetscited by 250
  8. 2023MimicPlay: Long-Horizon Imitation Learning by Watching Human Playcited by 223
  9. 2023Any-point Trajectory Modeling for Policy Learningcited by 215
  10. 2020Spatial-Temporal Transformer Networks for Traffic Flow Forecastingcited by 180
  11. 2024EgoMimic: Scaling Imitation Learning via Egocentric Videocited by 175
  12. 2018One-Shot Imitation from Observing Humans via Domain-Adaptive Meta-Learningcited by 148
  13. 2024Flow as the Cross-Domain Manipulation Interfacecited by 132
  14. 2023Learning to Act without Actionscited by 121
  15. 2023XSkill: Cross Embodiment Skill Discoverycited by 75
  16. 2025Motion Tracks: A Unified Representation for Human-Robot Transfer in Few-Shot Imitation Learningcited by 59
  17. 2024Vid2Robot: End-to-end Video-conditioned Policy Learning with Cross-Attention Transformerscited by 52
  18. 2022Cross-Domain Transfer via Semantic Skill Imitationcited by 18
  19. 2023Learning Video-Conditioned Policies for Unseen Manipulation Taskscited by 16

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 7; 2026: 28.

35 papers.

Sort:

By area:

  1. 2026π0.7: a Steerable Generalist Robotic Foundation Model with Emergent Capabilitiescited by 138
  2. 2025Vision-Language-Action Models for Robotics: A Review Towards Real-World Applicationscited by 84
  3. 2025Robot Learning from a Physical World Modelcited by 20
  4. 2026HumanEgo: Zero-Shot Robot Learning from Minutes of Human Egocentric Videoscited by 16
  5. 2025TraceGen: World Modeling in 3D Trace Space Enables Learning from Cross-Embodiment Videoscited by 15
  6. 2026ConLA: Contrastive Latent Action Learning from Human Videos for Robotic Manipulationcited by 14
  7. 2026MVP-LAM: Learning Action-Centric Latent Action via Cross-Viewpoint Reconstructioncited by 14
  8. 2025Co-Evolving Latent Action World Modelscited by 13
  9. 2025Developing Vision-Language-Action Model from Egocentric Videoscited by 11
  10. 2025OSMO: Open-Source Tactile Glove for Human-to-Robot Skill Transfercited by 10
  11. 2026Olaf-World: Orienting Latent Actions for Video World Modelingcited by 10
  12. 2026DiLA: Disentangled Latent Action World Modelscited by 10
23 more
  1. 2026Cross-Hand Latent Representation for Vision-Language-Action Modelscited by 7
  2. 2026Hierarchical Latent Action Modelcited by 6
  3. 2026UniLACT: Depth-Aware RGB Latent Action Learning for Vision-Language-Action Modelscited by 5
  4. 2026Robot Learning from Human Videos: A Surveycited by 5
  5. 2026HARP-VLA: Human-Robot Aligned Representation Learning for Vision-Language-Action Modelcited by 5
  6. 2026Human-Centric Transferable Tactile Pre-Training for Dexterous Robotic Manipulationcited by 5
  7. 2025Learning Skills from Action-Free Videoscited by 3
  8. 2026SeeTraceAct: Visibility-Aware Latent Planning from Cross-Embodiment Demonstration Videoscited by 3
  9. 2026Imitation from Heterogeneous Demonstrations using Grounded Latent-Action World Modelscited by 2
  10. 2026ShadowDancer: Teaching Video World Models Any Action by Learning Unified Dynamics Representations from a Video and Its Shadowcited by 2
  11. 2026The Imitator Game: Benchmarking Robot Imitative Ability Beyond Action Predictioncited by 1
  12. 2026PonderPounce: A Pretrained MLLM as an Episode Context Engine for Robot Controlcited by 1
  13. 2026In-Context Learning for Robots: Methods and Applicationscited by 1
  14. 2026EgoRecovery: Acquiring Failure Recovery Ability Through Human Recovery Demonstrationcited by 0
  15. 2026BooST: Bridging Semantics and Motions for Efficient Skill Transfercited by 0
  16. 2026ZETA: A Controlled Study of Zero-Shot Cross-Embodiment VLA Transfer for Tabletop Manipulationcited by 0
  17. 2026Toward Unified Robot Learning: Bridging Representation, Vision-Language-Action, and World Modelscited by 0
  18. 2026Reconstructing Is Not Acting: Action-Centric Latent Dynamics Modelingcited by 0
  19. 2026GALA: Geometry-Aware Latent Action Modeling for Vision-Language-Action Model Pretraining across Embodimentscited by 0
  20. 2026Video2Skill: From Streaming Experience to Reusable Embodied Skillscited by 0
  21. 2026Learning Transferable World-Action Models from Task-Paired Human Videoscited by 0
  22. 2026UniLAP: A Unified Latent Action Policy for Cross-Embodiment Manipulationcited by 0
  23. 2026DexStitch: Learning and Composing Dexterous Skills from Human Videos via Contact-Aware Robotic Inpaintingcited by 0