Corpus ID: 58981832

Visual Imitation Learning with Recurrent Siamese Networks

  title={Visual Imitation Learning with Recurrent Siamese Networks},
  author={G. Berseth and C. Pal},
  • G. Berseth, C. Pal
  • Published 2019
  • Computer Science, Mathematics
  • ArXiv
  • It would be desirable for a reinforcement learning (RL) based agent to learn behaviour by merely watching a demonstration. However, defining rewards that facilitate this goal within the RL paradigm remains a challenge. Here we address this problem with Siamese networks, trained to compute distances between observed behaviours and the agent's behaviours. Given a desired motion such Siamese networks can be used to provide a reward signal to an RL agent via the distance between the desired motion… CONTINUE READING

    Figures and Topics from this paper.


    Publications referenced by this paper.
    Learning Actionable Representations from Visual Observations
    • 22
    • PDF
    Visual Reinforcement Learning with Imagined Goals
    • 150
    • PDF
    Robust Imitation of Diverse Behaviors
    • 85
    • PDF
    Apprenticeship learning via inverse reinforcement learning
    • 1,933
    • Highly Influential
    • PDF
    Continuous control with deep reinforcement learning
    • 3,528
    • Highly Influential
    • PDF
    Imitation from Observation: Learning to Imitate Behaviors from Raw Video via Context Translation
    • 131
    • PDF
    One-Shot Visual Imitation Learning via Meta-Learning
    • 212
    • PDF
    Behavioral Cloning from Observation
    • 124
    • PDF
    Generative Adversarial Imitation Learning
    • 826
    • Highly Influential
    • PDF
    InfoGAIL: Interpretable Imitation Learning from Visual Demonstrations
    • 155
    • PDF