game? game exposes their API?
Reinforced learning agent? It doesn't necessarily have to expose the api. You can just get the specific frames of the game and feed it as data. As for button presses you can get it from the platform where your game runs
how about determining reward/penalty action? how to you get such data from the game to determine your agent actions?
If it's not a heavy game, you could just use object detection to get the score on the frame, the player position etc. I wouldn't do that if you have access to the game api.
Обсуждают сегодня