Enabling Visual Composition and Animation in Unsupervised Video Generation

arXiv:2403.14368v1 Announce Type: new
Abstract: In this work we propose a novel method for unsupervised controllable video generation. Once trained on a dataset of unannotated videos, at inference our model is capable of both composing scenes of predefined object parts and animating them in a plausible and controlled way. This is achieved by conditioning video generation on a randomly selected subset of local pre-trained self-supervised features during training. We call our model CAGE for visual Composition and Animation for video …

animation arxiv cs.cv enabling type unsupervised video video generation visual