Can you train a deep learning agent in video game and then transfer this learned information to a real world robot?
How can we handle the differences in visuals and action spaces between the game and reality?
These questions are researched in our conference article, “From Video Game to Real Robot: The Transfer Between Action Spaces”, which was kindly accepted to ICASSP 2020 conference. The research work was conducted while I was working at University of Eastern Finland and is continuation for my master’s thesis.
Check our conference presentation below!
