This paper reviews the NTIRE 2026 challenge on efficient single-image super-resolution with a focus on the proposed solutions and results. The aim of this challenge is to devise a network that reduces one or several aspects, such as runtime, parameters, and FLOPs, while maintaining PSNR of around 26.90 dB on the DIV2K_LSDIR_valid dataset, and 26.99 dB on the DIV2K_LSDIR_test dataset. The challenge had 95 registered participants, and 15 teams made valid submissions. They gauge the state-of-the-art results for efficient single-image super-resolution.
@article{arxiv.2604.03198,
title = {The Eleventh NTIRE 2026 Efficient Super-Resolution Challenge Report},
author = {Bin Ren and Hang Guo and Yan Shu and Jiaqi Ma and Ziteng Cui and Shuhong Liu and Guofeng Mei and Lei Sun and Zongwei Wu and Fahad Shahbaz Khan and Salman Khan and Radu Timofte and Yawei Li and Hongyuan Yu and Pufan Xu and Chen Wu and Long Peng and Jiaojiao Yi and Siyang Yi and Yuning Cui and Jingyuan Xia and Xing Mou and Keji He and Jinlin Wu and Zongang Gao and Sen Yang and Rui Zheng and Fengguo Li and Yecheng Lei and Wenkai Min and Jie Liu and Keye Cao and Shubham Sharma and Manish Prasad and Haobo Li and Matin Fazel and Abdelhak Bentaleb and Rui Chen and Shurui Shi and Zitao Dai and Qingliang Liu and Yang Cheng and Jing Hu and Xuan Zhang and Rui Ding and Tingyi Zhang and Hui Deng and Mengyang Wang and Fulin Liu and Jing Wei and Qian Wang and Hongying Liu and Mingyang Li and Guanglu Dong and Zheng Yang and Chao Ren and Hongbo Fang and Lingxuan Li and Lin Si and Pan Gao and Moncef Gabbouj and Watchara Ruangsang and Supavadee Aramvith},
journal= {arXiv preprint arXiv:2604.03198},
year = {2026}
}
Comments
CVPR 2026 NTIRE Workshop Paper, Efficient Super Resolution Technical Report