Asynchronous Methods for Deep Reinforcement Learning: Labyrinth
Channel:
Subscribers:
663,000
Published on ● Video Link: https://www.youtube.com/watch?v=nMR5mjCFZCw
The video shows an agent collecting rewards in previously unseen mazes using only raw pixels as input. The agent was trained using the Asynchronous Advantage Actor-Critic (A3C) algorithm and was only rewarded for picking up apples and orange portals during training.
Paper link - http://arxiv.org/pdf/1602.01783.pdf
Other Videos By Google DeepMind
Other Statistics
Minecraft Statistics For Google DeepMind
At this time, Google DeepMind has 128,442 views for Minecraft spread across 1 video. Less than an hour worth of Minecraft videos were uploaded to his channel, less than 0.01% of the total video content that Google DeepMind has uploaded to YouTube.