ildmxGL / qwop

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Play QWOP game using reinforcement learning

CS496 2018-2019 Winter Madcamp 4th Week

play game at here: http://www.foddy.net/Athletics.html. Screenshot using selenium and train the action with dqn. Press button is applied using window api. Getting score using tesseract-ocr. We downloaded it by referring to this site: http://egloos.zum.com/mcchae/v/11286830. Although he trained for more than 30 hours, 13 meters is the best score. AI can walk three meters long.

Reference

  1. https://blog.paperspace.com/dino-run/

About


Languages

Language:Jupyter Notebook 97.8%Language:Python 2.2%