This work from Berkeley + Google Brain describes using RL and GANs for transferring motion styles from animals successfully onto robots.
Super neat idea, and love to see ML being used more and more on real robots!
Love the Cost of Transport analysis - naturalistic movement really is very efficient, and good luck getting RL to solve the hard exploration problem of good motion and task performance simultaneously tabula rasa!
It's great to see hand engineered reward functions go the way of the dodo[2] - people are finally coming around to the idea that it was just feature engineering in a difficult-to-reason-about (reward) space! Down with hand specified reward functions! Let imitating nature reign supreme.
What's next, GANs for moral style transfer?[1]
[1] https://imgflip.com/i/6asa47
