G3AN: Disentangling Appearance and Motion for Video Generation

TL;DR

G3AN is a novel spatio-temporal generative model that disentangles appearance and motion to generate realistic human videos, outperforming existing methods on multiple datasets.

Contribution

The paper introduces G3AN, a three-stream generator that effectively models and disentangles appearance and motion in video generation.

Findings

Outperforms state-of-the-art on facial expression datasets

Achieves significant improvements on human action datasets

Successfully learns disentangled appearance and motion representations

Abstract

Creating realistic human videos entails the challenge of being able to simultaneously generate both appearance, as well as motion. To tackle this challenge, we introduce G$^{3}$AN, a novel spatio-temporal generative model, which seeks to capture the distribution of high dimensional video data and to model appearance and motion in disentangled manner. The latter is achieved by decomposing appearance and motion in a three-stream Generator, where the main stream aims to model spatio-temporal consistency, whereas the two auxiliary streams augment the main stream with multi-scale appearance and motion features, respectively. An extensive quantitative and qualitative analysis shows that our model systematically and significantly outperforms state-of-the-art methods on the facial expression datasets MUG and UvA-NEMO, as well as the Weizmann and UCF101 datasets on human action. Additional analysis on the learned latent representations confirms the successful decomposition of appearance and motion. Source code and pre-trained models are publicly available.