Intel AI researchers combine reinforcement learning methods to teach 3D humanoid how to walk

Intel AI researchers have devised methods to combine forms of reinforcement learning in order to teach OpenAI's Humanoid how to walk without falling.