freeCodeCamp.org
Hear the story of how we used Python and machine learning to build an artificial intelligence that plays Super StreetFighter II on the Super NES. We’ll cover how Python provided the key glue between the SNES emulator and AI, and how the AI was built with gym, keras-rl and tensorflow. We’ll show examples of game play and training, and talk about which bot beat which bot in the bot-v-bot tournament we ran.
Talk given by Adam Fletcher and Jonathan Mortensen at PyCon 2018.
Thanks to PyCon for giving us permission to post this talk. freeCodeCamp is not associated with this talk. We’re just excited to bring more exposure to to it!
—
Learn to code for free and get a developer job: https://www.freecodecamp.com
Read hundreds of articles on programming: https://medium.freecodecamp.com
And subscribe for new videos on technology every day: https://youtube.com/subscription_center?add_user=freecodecamp
Source
A revolution without dancing is a revolution not worth having.
Awesome use of machine learning! If you like AI applied to games you might want to give my channel a check. Cheers!
Came here to see a match. 🙁
wow! so +++++ interesting
great questions from audiences
Would have liked to hear more about the observation space. From the talk we heard x position and health bars are observed. I wonder what else about the opponent's action is known to the agent.
Would it be possible to do the same approach to other snes games?
Do it for fortnite
whoa what, the presentation just ended when I thought it was just getting warmed up..
where was the main model's architecture shown? you talked about some boring infrastructure setup stuff and about your cute on site display but not the main part??
wrg, big doesnt matr