QT-Opt: Scalable Deep Reinforcement Learning for Vision-Based Robotic Manipulation

For more information, see http://ai.googleblog.com/2018/06/scalable-deep-reinforcement-learning.html.

source