arXiv 1912.01603
Dream to Control: Learning Behaviors by Latent Imagination
By Danijar Hafner, Timothy Lillicrap, et al.
Published 2019-12-03
Citation lineage
Review the prior work and downstream research connected to this paper.
Learned world models summarize an agent's experience to facilitate learning complex behaviors. While learning world models from high-dimensional sensory inputs is becoming feasible through deep learning, there are many potential ways for deriving behaviors from them. We present Dreamer, a reinforcement learning agent that solves long-horizon tasks from images purely by latent imagination. We efficiently learn behavi…