Reinforcement learning models in ViZDoom environment
Switch branches/tags
Nothing to show
Clone or download
Latest commit cd27adf Oct 22, 2018
Permalink
Failed to load latest commit information.
actions missing action sets Oct 21, 2018
checkpoints refactoring Nov 7, 2017
environments add research projects Oct 20, 2018
experiments add research projects Oct 20, 2018
images bt link Oct 22, 2018
src basic behavior tree Oct 21, 2018
tools add research projects Oct 20, 2018
trained_models new experiments Aug 20, 2017
LICENSE.md Update LICENSE.md Oct 21, 2018
README.md Update README.md Oct 22, 2018

README.md

DoomNet

DoomNet is a ViZDoom agent trained by reinforcement learning. The agent is a neural network that outputs a probability of actions given only pixels from the screen buffer and set of game variables.
DoomNet is a 1st Runner-Up at Visual Doom AI Competition 2018.

What a simple behavior tree can do

DoomNet on a simple behavior tree

1st Runner-Up at Visual Doom AI Competition 2018

DoomNet track1, submission 0

Visual Doom AI Competition 2017

Joint work with Bobby DeSimone

DoomNet's view is at left in the middle row
DoomNet track1, elimination round 2017

D3 Battle

DoomNet trained on D3-battle config

Health Gathering

DoomNet trained on health gathering config

Rocket Basic

DoomNet trained on basic rocket config

Basic

DoomNet trained on basic config