Description
Interest group for ViZDoom framework and Visual Doom AI competitions.
ViZDoom allows developing AI bots that play Doom using only the visual information (the screen buffer). It is primarily intended for research in machine visual learning, and reinforcement deep learning, in particular.