ImaGINator: Conditional Spatio-Temporal GAN for Video Generation

Yaohui Wang, Piotr Biliński, François Brémond, Antitza Dantcheva

2020 · 109 citations · 41 references

DOIFull text

Open access

Concepts

TL;DR

Generating human videos from a single image requires simultaneously realistic appearance and motion. The paper proposes ImaGINator, a conditional GAN that decomposes appearance and motion to generate realistic videos from a single image and a condition. ImaGINator uses a spatio‑temporal fusion scheme that preserves appearance across layers and a factorized transposed (1+2)D convolution for efficient, high‑quality video synthesis, evaluated on MUG, UvA‑NEMO, NATOPS, and Weizmann. The approach achieves significantly better quantitative and qualitative results than the state‑of‑the‑art. Source code and models are available at https://github.com/wyhsirius/ImaGINator.

Abstract

Generating human videos based on single images entails the challenging simultaneous generation of realistic and visual appealing appearance and motion. In this context, we propose a novel conditional GAN architecture, namely ImaGINator, which given a single image, a condition (label of a facial expression or action) and noise, decomposes appearance and motion in both latent and high level feature spaces, generating realistic videos. This is achieved by (i) a novel spatio-temporal fusion scheme, which generates dynamic motion, while retaining appearance throughout the full video sequence by transmitting appearance (originating from the single image) through all layers of the network. In addition, we propose (ii) a novel transposed (1+2)D convolution, factorizing the transposed 3D convolutional filters into separate transposed temporal and spatial components, which yields significantly gains in video quality and speed. We extensively evaluate our approach on the facial expression datasets MUG and UvA-NEMO, as well as on the action datasets NATOPS and Weizmann. We show that our approach achieves significantly better quantitative and qualitative results than the state-of-the-art. The source code and models are available under https://github.com/wyhsirius/ImaGINator.

References

41