mlsanityData to 7 Oct 2026

Back to Reality for Imitation Learning

arXiv 2111.12867 ↗First posted 25 Nov 2021cs.RO cs.LGcited by 3

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

Abstract (arXiv)

Imitation learning, and robot learning in general, emerged due to breakthroughs in machine learning, rather than breakthroughs in robotics. As such, evaluation metrics for robot learning are deeply rooted in those for machine learning, and focus primarily on data efficiency. We believe that a better metric for real-world robot learning is time efficiency, which better models the true cost to humans. This is a call to arms to the robot learning community to develop our own evaluation metrics, tailored towards the long-term goals of real-world robotics.

Key facts. Cited by 3 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 19 of its 27 references that are in the map. First posted on arXiv 25 Nov 2021.

Builds on

19 of 27 references.

Closest ancestors · 3

Same direction, newest first

  1. 2021Learning Multi-Stage Tasks with One Demonstration via Self-Replay14 Nov
  2. 2021Coarse-to-Fine Imitation Learning: Robot Manipulation from a Single Demonstration13 May
  3. 2020Visual Imitation Made Easy11 Aug

Other references · 16

  1. 2018An Algorithmic Perspective on Imitation Learningcited by 455
  2. 2020Learning from Suboptimal Demonstration via Self-Supervised Reward Regressioncited by 84
  3. 2020Transformers for One-Shot Visual Imitationcited by 77
  4. 2020f-IRL: Inverse Reinforcement Learning via State Marginal Matchingcited by 67
  5. 2019Deep Reinforcement Learning for Industrial Insertion Tasks with Visual Inputs and Natural Rewardscited by 58
  6. 2020Model-Based Inverse Reinforcement Learning from Visual Demonstrationscited by 52
  7. 2019Task-Relevant Adversarial Imitation Learningcited by 45
  8. 2019Positive-Unlabeled Reward Learningcited by 36
  9. 2020Learning Dexterous Manipulation from Suboptimal Expertscited by 23
  10. 2021Learning from Demonstrations using Signal Temporal Logiccited by 17
  11. 2020Modeling Long-horizon Tasks as Sequential Interaction Landscapescited by 14
  12. 2020Learning Object Manipulation Skills via Approximate State Estimation from Real Videoscited by 12
4 more
  1. 2020Interactive Imitation Learning in State-Spacecited by 10
  2. 2020CLOUD: Contrastive Learning of Unsupervised Dynamicscited by 4
  3. 2020Generative adversarial training of product of policies for robust and adaptive movement primitivescited by 2
  4. 2021Tolerance-Guided Policy Learning for Adaptable and Transferrable Delicate Industrial Insertioncited by 2

Built on by

Cited by year

Citing papers per year: 2024: 2; 2025: 1.

3 papers.

  1. 2024MILES: Making Imitation Learning Easy with Self-Supervisioncited by 9
  2. 2024"Give Me an Example Like This": Episodic Active Reinforcement Learning from Demonstrationscited by 2
  3. 2025ASkDAgger: Active Skill-level Data Aggregation for Interactive Imitation Learningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026An Efficient Metric for Data Quality Measurement in Imitation Learning
  2. 2024Data Scaling Laws in Imitation Learning for Robotic Manipulation
  3. 2020Scalable Multi-Task Imitation Learning with Autonomous Improvement
  4. 2019Efficient Supervision for Robot Learning via Imitation, Simulation, and Adaptation
  5. 2025Learning a Thousand Tasks in a Day
  6. 2024Generalized Robot Learning Framework
  7. 2023Data Quality in Imitation Learning
  8. 2022Imitation Learning by Estimating Expertise of Demonstrators