Learned Compression Artifact Removal by Deep Residual Networks

Ogun Kirmemis, Gonca Bakar, A. Murat Tekalp; The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2018, pp. 2602-2605

Abstract


We propose a method for learned compression artifact removal by post-processing of BPG compressed images. We trained three networks of different sizes. We encoded input images using BPG with different QP values. We submitted the best combination of test images, encoded with different QP and post-processed by one of three networks, which satisfy the file size and decode time constraints imposed by the Challenge. The selection of the best combination is posed as an integer programming problem. Although the visual improvements in image quality is impressive, the average PSNR improvement for the results is about 0.5 dB.

Related Material


[pdf] [arXiv]
[bibtex]
@InProceedings{Kirmemis_2018_CVPR_Workshops,
author = {Kirmemis, Ogun and Bakar, Gonca and Murat Tekalp, A.},
title = {Learned Compression Artifact Removal by Deep Residual Networks},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops},
month = {June},
year = {2018}
}