mlsanityData to 7 Oct 2026

Versatile Skill Control via Self-supervised Adversarial Imitation of Unlabeled Mixed Motions

arXiv 2209.07899 ↗First posted 16 Sep 2022cs.RO cs.AI cs.LGcited by 13

On the map: Offline RL and imitation learning from data → Imitation learning methods

Abstract (arXiv)

Learning diverse skills is one of the main challenges in robotics. To this end, imitation learning approaches have achieved impressive results. These methods require explicitly labeled datasets or assume consistent skill execution to enable learning and active control of individual behaviors, which limits their applicability. In this work, we propose a cooperative adversarial method for obtaining single versatile policies with controllable skill sets from unlabeled datasets containing diverse state transition patterns by maximizing their discriminability. Moreover, we show that by utilizing unsupervised skill discovery in the generative adversarial imitation learning framework, novel and useful skills emerge with successful task fulfillment. Finally, the obtained versatile policies are tested on an agile quadruped robot called Solo 8 and present faithful replications of diverse skills encoded in the demonstrations.

Key facts. Cited by 13 papers in the map (data through 7 Oct 2026). Placed in Imitation learning methods, one of 13 directions in Offline RL and imitation learning from data. Builds on 24 of its 32 references that are in the map. First posted on arXiv 16 Sep 2022.

Builds on

24 of 32 references.

References · 24

  1. 2017Proximal Policy Optimization Algorithmscited by 14,441
  2. 2016Generative Adversarial Imitation Learningcited by 1,576
  3. 2016Least Squares Generative Adversarial Networkscited by 1,000
  4. 2021Isaac Gym: High Performance GPU-Based Physics Simulation For Robot Learningcited by 882
  5. 2018Diversity is All You Need: Learning Skills without a Reward Functioncited by 734
  6. 2020Learning Quadrupedal Locomotion over Challenging Terraincited by 494
  7. 2021RMA: Rapid Motor Adaptation for Legged Robotscited by 430
  8. 2019Learning agile and dynamic motor skills for legged robotscited by 402
  9. 2019Dynamics-Aware Unsupervised Discovery of Skillscited by 308
  10. 2022Learning robust perceptive locomotion for quadrupedal robots in the wildcited by 299
  11. 2016Variational Intrinsic Controlcited by 262
  12. 2019MAVEN: Multi-Agent Variational Explorationcited by 181
12 more
  1. 2015Variational Information Maximisation for Intrinsically Motivated Reinforcement Learningcited by 142
  2. 2019Fast Task Inference with Variational Intrinsic Successor Featurescited by 131
  3. 2017Stochastic Neural Networks for Hierarchical Reinforcement Learningcited by 124
  4. 2018Variational Option Discovery Algorithmscited by 121
  5. 2020Explore, Discover and Learn: Unsupervised Discovery of State-Covering Skillscited by 120
  6. 2021Reward is enough for convex MDPscited by 82
  7. 2020One Solution is Not All You Need: Few-Shot Extrapolation via Structured MaxEnt RLcited by 79
  8. 2021Learning more skills through optimistic explorationcited by 37
  9. 2020Relative Variational Intrinsic Controlcited by 34
  10. 2022Learning Agile Skills via Adversarial Imitation of Rough Partial Demonstrationscited by 30
  11. 2022Advanced Skills through Multiple Adversarial Motion Priors in Reinforcement Learningcited by 23
  12. 2019An Open Torque-Controlled Modular Robot Architecture for Legged Locomotion Researchcited by 18

Built on by

Cited by year

Citing papers per year: 2023: 5; 2024: 3; 2025: 5.

13 papers.

Sort:

By area:

  1. 2025Robotic World Model: A Neural Network Simulator for Robust Policy Optimization in Roboticscited by 47
  2. 2023Prompt, Plan, Perform: LLM-based Humanoid Control via Quantized Imitation Learningcited by 31
  3. 2023RL + Model-based Control: Using On-demand Optimal Control to Learn Versatile Legged Locomotioncited by 14
  4. 2024FLD: Fourier Latent Dynamics for Structured Motion Representation and Learningcited by 14
  5. 2023Controlling the Solo12 Quadruped Robot with Deep Reinforcement Learningcited by 11
  6. 2024Rethinking Robustness Assessment: Adversarial Attacks on Learning-based Quadrupedal Locomotion Controllerscited by 9
  7. 2024Guided Reinforcement Learning for Robust Multi-Contact Loco-Manipulationcited by 9
  8. 2023Learning Diverse Skills for Local Navigation under Multi-constraint Optimalitycited by 7
  9. 2023Fully Spiking Neural Network for Legged Robotscited by 4
  10. 2025Symmetry-Guided Memory Augmentation for Efficient Locomotion Learningcited by 3
  11. 2025Integrating Diffusion-based Multi-task Learning with Online Reinforcement Learning for Robust Quadruped Robot Controlcited by 3
  12. 2025Feature-Based vs. GAN-Based Learning from Demonstrations: When and Whycited by 1
1 more
  1. 2025Unlocking the Potential of Soft Actor-Critic for Imitation Learningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Continuous Control of Diverse Skills in Quadruped Robots Without Complete Expert Datasets
  2. 2020ADAIL: Adaptive Adversarial Imitation Learning
  3. 2026Generative adversarial imitation learning for robot swarms: Learning from human demonstrations and trained policies
  4. 2022Latent Policies for Adversarial Imitation Learning
  5. 2026Diverse Skill Discovery for Quadruped Robots via Unsupervised Learning
  6. 2023Unlabeled Imperfect Demonstrations in Adversarial Imitation Learning
  7. 2023Offline Diversity Maximization Under Imitation Constraints
  8. 2023Sample-efficient Adversarial Imitation Learning