English

Video Generation from Single Semantic Label Map

Computer Vision and Pattern Recognition 2019-03-12 v1

Abstract

This paper proposes the novel task of video generation conditioned on a SINGLE semantic label map, which provides a good balance between flexibility and quality in the generation process. Different from typical end-to-end approaches, which model both scene content and dynamics in a single step, we propose to decompose this difficult task into two sub-problems. As current image generation methods do better than video generation in terms of detail, we synthesize high quality content by only generating the first frame. Then we animate the scene based on its semantic meaning to obtain the temporally coherent video, giving us excellent results overall. We employ a cVAE for predicting optical flow as a beneficial intermediate step to generate a video sequence conditioned on the initial single frame. A semantic label map is integrated into the flow prediction module to achieve major improvements in the image-to-video generation process. Extensive experiments on the Cityscapes dataset show that our method outperforms all competing methods.

Keywords

Cite

@article{arxiv.1903.04480,
  title  = {Video Generation from Single Semantic Label Map},
  author = {Junting Pan and Chengyu Wang and Xu Jia and Jing Shao and Lu Sheng and Junjie Yan and Xiaogang Wang},
  journal= {arXiv preprint arXiv:1903.04480},
  year   = {2019}
}

Comments

Paper accepted at CVPR 2019. Source code and models available at https://github.com/junting/seg2vid/tree/master

R2 v1 2026-06-23T08:04:38.720Z