NTIRE 2022 Burst Super-Resolution Challenge

Goutam Bhat, Martin Danelljan, Radu Timofte, Yizhen Cao, Yuntian Cao, Meiya Chen, Xihao Chen, Shen Cheng, Akshay Dudhane, Haoqiang Fan, Ruipeng Gang, Jian Gao, Yan Gu, Jie Huang, Liufeng Huang, Youngsu Jo, Sukju Kang, Salman Khan, Fahad Shahbaz Khan, Yuki Kondo, Chenghua Li, Fangya Li, Jinjing Li, Youwei Li, Zechao Li, Chenming Liu, Shuaicheng Liu, Zikun Liu, Zhuoming Liu, Ziwei Luo, Zhengxiong Luo, Nancy Mehta, Subrahmanyam Murala, Yoonchan Nam, Chihiro Nakatani, Pavel Ostyakov, Jinshan Pan, Ge Song, Jian Sun, Long Sun, Jinhui Tang, Norimichi Ukita, Zhihong Wen, Qi Wu, Xiaohe Wu, Zeyu Xiao, Zhiwei Xiong, Rongjian Xu, Ruikang Xu, Youliang Yan, Jialin Yang, Wentao Yang, Zhongbao Yang, Fuma Yasue, Mingde Yao, Lei Yu, Cong Zhang, Syed Waqas Zamir, Jianxing Zhang, Shuohao Zhang, Zhilu Zhang, Qian Zheng, Gaofeng Zhou, Magauiya Zhussip, Xueyi Zou, Wangmeng Zuo; Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2022, pp. 1041-1061

Abstract

Burst super-resolution has received increased attention in recent years due to its applications in mobile photography. By merging information from multiple shifted images of a scene, burst super-resolution aims to recover details which otherwise cannot be obtained using a simple input image. This paper reviews the NTIRE 2022 challenge on burst super-resolution. In the challenge, the participants were tasked with generating a clean RGB image with 4xhigher resolution, given a RAW noisy burst as input. That is, the methods need to perform joint denoising, demosaicking, and super-resolution. The challenge consisted of 2 tracks. Track 1 employed synthetic data, where pixel-accurate high-resolution ground truths are available. Track 2 on the other hand used real-world bursts captured from a handheld camera, along with approximately aligned reference images captured using a DSLR. 14 teams participated in the final testing phase. The top performing methods establish a new state-of-the-art on the burst super-resolution task.

Related Material

[pdf]
[bibtex]
@InProceedings{Bhat_2022_CVPR, author = {Bhat, Goutam and Danelljan, Martin and Timofte, Radu and Cao, Yizhen and Cao, Yuntian and Chen, Meiya and Chen, Xihao and Cheng, Shen and Dudhane, Akshay and Fan, Haoqiang and Gang, Ruipeng and Gao, Jian and Gu, Yan and Huang, Jie and Huang, Liufeng and Jo, Youngsu and Kang, Sukju and Khan, Salman and Khan, Fahad Shahbaz and Kondo, Yuki and Li, Chenghua and Li, Fangya and Li, Jinjing and Li, Youwei and Li, Zechao and Liu, Chenming and Liu, Shuaicheng and Liu, Zikun and Liu, Zhuoming and Luo, Ziwei and Luo, Zhengxiong and Mehta, Nancy and Murala, Subrahmanyam and Nam, Yoonchan and Nakatani, Chihiro and Ostyakov, Pavel and Pan, Jinshan and Song, Ge and Sun, Jian and Sun, Long and Tang, Jinhui and Ukita, Norimichi and Wen, Zhihong and Wu, Qi and Wu, Xiaohe and Xiao, Zeyu and Xiong, Zhiwei and Xu, Rongjian and Xu, Ruikang and Yan, Youliang and Yang, Jialin and Yang, Wentao and Yang, Zhongbao and Yasue, Fuma and Yao, Mingde and Yu, Lei and Zhang, Cong and Zamir, Syed Waqas and Zhang, Jianxing and Zhang, Shuohao and Zhang, Zhilu and Zheng, Qian and Zhou, Gaofeng and Zhussip, Magauiya and Zou, Xueyi and Zuo, Wangmeng}, title = {NTIRE 2022 Burst Super-Resolution Challenge}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops}, month = {June}, year = {2022}, pages = {1041-1061} }