Learned Video Compression with Feature-Level Residuals
Abstract
In this paper, we present an end-to-end video compression network for P-frame challenge on CLIC. We focus on deep neural network (DNN) based video compression, and improve the current frameworks from three aspects. First, we notice that pixel space residuals is sensitive to the prediction errors of optical flow based motion compensation. To suppress the relative influence, we propose to compress the residuals of image feature rather than the residuals of image pixels. Furthermore, we combine the advantages of both pixel-level and feature-level residual compression methods by model ensembling. Finally, we propose a step-by-step training strategy to improve the training efficiency of the whole framework. Experiment results indicate that our proposed method achieves 0.9968 MS-SSIM on CLIC validation set and 0.9967 MS-SSIM on test set.
Cite
Text
Feng et al. "Learned Video Compression with Feature-Level Residuals." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2020. doi:10.1109/CVPRW50498.2020.00068Markdown
[Feng et al. "Learned Video Compression with Feature-Level Residuals." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2020.](https://mlanthology.org/cvprw/2020/feng2020cvprw-learned/) doi:10.1109/CVPRW50498.2020.00068BibTeX
@inproceedings{feng2020cvprw-learned,
title = {{Learned Video Compression with Feature-Level Residuals}},
author = {Feng, Runsen and Wu, Yaojun and Guo, Zongyu and Zhang, Zhizheng and Chen, Zhibo},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
year = {2020},
pages = {529-532},
doi = {10.1109/CVPRW50498.2020.00068},
url = {https://mlanthology.org/cvprw/2020/feng2020cvprw-learned/}
}