-
[pdf]
[arXiv]
[bibtex]@InProceedings{Gushchin_2026_CVPR, author = {Gushchin, Aleksandr and Abud, Khaled and Shumitskaya, Ekaterina and Filippov, Artem and Bychkov, Georgii and Lavrushkin, Sergey and Erofeev, Mikhail and Antsiferova, Anastasia and Chen, Changsheng and Tan, Shunquan and Timofte, Radu and Vatolin, Dmitriy and Song, Chuanbiao and Yu, Zijian and Tan, Hao and Lan, Jun and Yang, Zhiqiang and Tang, Yongwei and Wu, Zhiqiang and Seow, Jia Wen and Koay, Hong Vin and Ren, Haodong and Xu, Feng and Chen, Shuai and Xia, Ruiyang and Zhang, Qi and Xu, Yaowen and Zou, Zhaofan and Sun, Hao and Lu, Dagong and Yao, Mufeng and Xu, Xinlei and Wu, Fei and Guo, Fengjun and Luo, Cong and Sharma, Hardik and Negi, Aashish and Shaily, Prateek and Kumar, Jayant and Chaudhary, Sachin and Dudhane, Akshay and Hambarde, Praful and Shukla, Amit and Tu, Zhilin and Li, Fengpeng and Zhang, Jiamin and Fei, Jianwei and Li, Kemou and Wu, Haiwei and Benjdira, Bilel and Ali, Anas M. and Boulila, Wadii and Qu, Chenfan and Li, Junchi}, title = {NTIRE 2026 Challenge on Robust AI-Generated Image Detection in the Wild}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops}, month = {June}, year = {2026}, pages = {1895-1913} }
NTIRE 2026 Challenge on Robust AI-Generated Image Detection in the Wild
Abstract
This paper presents an overview of the NTIRE 2026 Challenge on Robust AI-Generated Image Detection in the Wild, held in conjunction with the NTIRE workshop at CVPR 2026. The goal of this challenge was to develop detection models capable of distinguishing real images from generated ones in realistic scenarios: the images are often transformed (cropped, resized, compressed, blurred) for practical usage, and therefore, the detection models should be robust to such transformations. The challenge is based on a novel dataset consisting of 108,750 real and 185,750 AI-generated images from 42 generators comprising a large variety of open-source and closed-source models of various architectures, augmented with 36 image transformations. Methods were evaluated using ROC AUC on the full test set, including both transformed and untransformed images. A total of 511 participants registered, with 20 teams submitting valid final solutions. This report provides a comprehensive overview of the challenge, describes the proposed solutions, and can be used as a valuable reference for researchers and practitioners in increasing the robustness of the detection models to real-world transformations.
Related Material