MIPI 2024 少样本原始图像去噪挑战赛：方法与结果
计算机视觉与模式识别
2024-06-12 v1
摘要
移动平台上计算摄影和成像需求的日益增长，推动了先进图像传感器与新颖算法在相机系统中的广泛开发和集成。然而，用于研究的高质量数据的稀缺性，以及业界和学术界深入交流观点的难得机会，制约了移动智能摄影与成像（MIPI）的发展。基于先前在ECCV 2022和CVPR 2023上举办的MIPI研讨会的成果，我们推出了第三届MIPI挑战赛，包括三个专注于新型图像传感器和成像算法的赛道。在本文中，我们总结并回顾了MIPI 2024的少样本原始图像去噪赛道。共有165名参与者成功注册，7支队伍在最终测试阶段提交了结果。本次挑战赛中开发的解决方案在少样本原始图像去噪方面达到了最先进的性能。有关本次挑战赛的更多详情和数据集的链接，请访问https://mipichallenge.org/MIPI2024。
引用
@article{arxiv.2406.07006,
title = {MIPI 2024 Challenge on Few-shot RAW Image Denoising: Methods and Results},
author = {Xin Jin and Chunle Guo and Xiaoming Li and Zongsheng Yue and Chongyi Li and Shangchen Zhou and Ruicheng Feng and Yuekun Dai and Peiqing Yang and Chen Change Loy and Ruoqi Li and Chang Liu and Ziyi Wang and Yao Du and Jingjing Yang and Long Bao and Heng Sun and Xiangyu Kong and Xiaoxia Xing and Jinlong Wu and Yuanyang Xue and Hyunhee Park and Sejun Song and Changho Kim and Jingfan Tan and Wenhan Luo and Zikun Liu and Mingde Qiao and Junjun Jiang and Kui Jiang and Yao Xiao and Chuyang Sun and Jinhui Hu and Weijian Ruan and Yubo Dong and Kai Chen and Hyejeong Jo and Jiahao Qin and Bingjie Han and Pinle Qin and Rui Chai and Pengyuan Wang},
journal= {arXiv preprint arXiv:2406.07006},
year = {2024}
}
备注
CVPR 2024 Mobile Intelligent Photography and Imaging (MIPI) Workshop--Few-shot RAWImage Denoising Challenge Report. Website: https://mipi-challenge.org/MIPI2024/