This paper reviews the second AIM learned ISP challenge and provides the description of the proposed solutions and results. The participating teams were solving a real-world RAW-to-RGB mapping problem, where to goal was to map the original low-quality RAW images captured by the Huawei P20 device to the same photos obtained with the Canon 5D DSLR camera. The considered task embraced a number of complex computer vision subtasks, such as image demosaicing, denoising, white balancing, color and contrast correction, demoireing, etc. The target metric used in this challenge combined fidelity scores (PSNR and SSIM) with solutions' perceptual results measured in a user study. The proposed solutions significantly improved the baseline results, defining the state-of-the-art for practical image signal processing pipeline modeling.
@article{arxiv.2011.04994,
title = {AIM 2020 Challenge on Learned Image Signal Processing Pipeline},
author = {Andrey Ignatov and Radu Timofte and Zhilu Zhang and Ming Liu and Haolin Wang and Wangmeng Zuo and Jiawei Zhang and Ruimao Zhang and Zhanglin Peng and Sijie Ren and Linhui Dai and Xiaohong Liu and Chengqi Li and Jun Chen and Yuichi Ito and Bhavya Vasudeva and Puneesh Deora and Umapada Pal and Zhenyu Guo and Yu Zhu and Tian Liang and Chenghua Li and Cong Leng and Zhihong Pan and Baopu Li and Byung-Hoon Kim and Joonyoung Song and Jong Chul Ye and JaeHyun Baek and Magauiya Zhussip and Yeskendir Koishekenov and Hwechul Cho Ye and Xin Liu and Xueying Hu and Jun Jiang and Jinwei Gu and Kai Li and Pengliang Tan and Bingxin Hou},
journal= {arXiv preprint arXiv:2011.04994},
year = {2020}
}
Comments
Published in ECCV 2020 Workshops (Advances in Image Manipulation), https://data.vision.ee.ethz.ch/cvl/aim20/