UBC Computer Science is playing around with AI to make video games even more realistic.
"A team from Electronic Arts (EA) and the University of British Columbia is using a technique called 'reinforcement learning', which is loosely inspired by the way animals learn in response to positive and negative feedback, to automatically animate humanoid characters," says an article in Wired Magazine.
With artificial intelligence research at the helm of reinforcement learning, sporting games like FIFA are able to move toward automating the animation process. This can potentially save companies millions of dollars and make the games more efficient and realistic. The current method, motion capture, limits outcomes to the actions recorded and also requires someone to write code to animate the characters.
Hung Yu Ling and Michiel Van de Panne of UBC Computer Science, and Fabio Zinno and George Cheng of Electronic Arts have produced a paper to present their work at the Siggraph 2020 conference in July.