Making computer animation more agile, acrobatic — and realistic
04/10/18 — EECS grad student Xue Bin “Jason” Peng and his colleagues have made a major advance in realistic computer animation, using deep reinforcement learning to create a virtual stuntman that mimics natural motions.