Deep Stereo Image Compression via Bi-Directional Coding

Jianjun Lei, Xiangrui Liu, Bo Peng, Dengchao Jin, Wanqing Li, Jingxiao Gu

CVPR 2022 pp. 19669-19678

doi:10.1109/CVPR52688.2022.01905 /cvpr/2022/lei2022cvpr-deep/

Abstract

Existing learning-based stereo compression methods usually adopt a unidirectional approach to encoding one image independently and the other image conditioned upon the first. This paper proposes a novel bi-directional coding-based end-to-end stereo image compression network (BCSIC-Net). BCSIC-Net consists of a novel bi-directional contextual transform module which performs nonlinear transform conditioned upon the inter-view context in a latent space to reduce inter-view redundancy, and a bi-directional conditional entropy model that employs inter-view correspondence as a conditional prior to improve coding efficiency. Experimental results on the InStereo2K and KITTI datasets demonstrate that the proposed BCSIC-Net can effectively reduce the inter-view redundancy and outperforms state-of-the-art methods.

PDF CVPR Semantic Scholar

Cite

Text

Lei et al. "Deep Stereo Image Compression via Bi-Directional Coding." Conference on Computer Vision and Pattern Recognition, 2022. doi:10.1109/CVPR52688.2022.01905

Markdown

[Lei et al. "Deep Stereo Image Compression via Bi-Directional Coding." Conference on Computer Vision and Pattern Recognition, 2022.](https://mlanthology.org/cvpr/2022/lei2022cvpr-deep/) doi:10.1109/CVPR52688.2022.01905

BibTeX

@inproceedings{lei2022cvpr-deep,
  title     = {{Deep Stereo Image Compression via Bi-Directional Coding}},
  author    = {Lei, Jianjun and Liu, Xiangrui and Peng, Bo and Jin, Dengchao and Li, Wanqing and Gu, Jingxiao},
  booktitle = {Conference on Computer Vision and Pattern Recognition},
  year      = {2022},
  pages     = {19669-19678},
  doi       = {10.1109/CVPR52688.2022.01905},
  url       = {https://mlanthology.org/cvpr/2022/lei2022cvpr-deep/}
}