1 paper
Anton Arapin, Aditya Singh, Yixin Wang +2
World models of visual information predict future image frames conditioned on given action sequences. This predicted scene evolution is driven by deterministic changes such as cont…