Meta-learning For Robotics

Meta-Learning for Robotics is a formal framework within data-efficient robot learning that enables robotic systems to rapidly adapt to new tasks by acquiring inductive biases from prior experience across related tasks, thereby distinguishing itself from general learning approaches through its focus on learning-to-learn mechanisms specifically tailored for robotic domains.

65

Publications

3.8K

Citations

203

Authors

70

Institutions

Publications per year

2017–2026

56

Authors

203

  1. CF
    Chelsea Finn

    University of California, Berkeley, United States

    7 publications · 517 citations

    7

    h-index

  2. SL
    Sergey Levine

    University of California, Berkeley, United States

    6 publications · 581 citations

    6

    h-index

  3. PA
    Pieter Abbeel

    University of California, Berkeley, United States

    5 publications · 1K citations

    5

    h-index

  4. JT
    Jie Tan

    Google (United States), United States

    3 publications · 156 citations

    3

    h-index

  5. JP
    Jan Peters

    Technische Universität Darmstadt, Germany

    2 publications · 226 citations

    2

    h-index

1–5 of 203

Venues

  1. arXiv (Cornell University)

    Repository · 15 publications · 913 citations

    12

    h-index

  2. IEEE Robotics and Automation Letters

    7 publications · 342 citations

    7

    h-index

  3. 2

    h-index

  4. Neural Information Processing Systems

    Conference · 2 publications · 304 citations

    2

    h-index

  5. Conference on Robot Learning

    Conference · 2 publications · 135 citations

    2

    h-index

1–5