1 paper
Angel Villar-Corrales, Ismail Wahdan, Sven Behnke
We propose a novel framework for the task of object-centric video prediction, i.e., extracting the compositional structure of a video sequence, as well as modeling objects dynamics…