Unsupervised Imitation Learning

Pierre Sermanet, Corey Lynch †, Kelvin Xu †, Jasmine Hsu, Sergey Levine
Google Brain
(† Google Brain Residency program g.co/brainresidency)


The general goal of this project is to give to robots the ability to learn to imitate humans without any supervision.
We present the following papers as steps towards of that goal.

1. Time-Contrastive Networks

2. Unsupervised Perceptual Rewards