× Predicting Ego-centric Video from human Actions (PEVA). Given previous video frames and an motion specifying a desired change in 3D pose, PEVA predicts the subsequent video ..
