Phase-Functioned Neural Networks For Character Control

This exciting research from Daniel Holden, (University of Edinburgh), Taku Komura, (University of Edinburgh) & Jun Saito, (Method Studios) presents a machine-learning approach to driving characters.

“The entire network is trained in an end-to-end fashion on a large dataset composed of locomotion such as walking, running, jumping, and climbing movements fitted into virtual environments. Our system can therefore automatically produce motions where the character adapts to different geometric environments such as walking and running over rough terrain, climbing over large rocks, jumping over obstacles, and crouching under low ceilings.”

Read the paper here.