3 code implementations • ICLR 2019 • Shani Gamrian, Yoav Goldberg
Despite the remarkable success of Deep RL in learning control policies from raw pixels, the resulting models do not generalize.
Image-to-Image Translation Imitation Learning +4