-
[pdf]
[arXiv]
[bibtex]@InProceedings{Han_2025_CVPR, author = {Han, Shuhao and Fan, Haotian and Kong, Fangyuan and Liao, Wenjie and Guo, Chunle and Li, Chongyi and Timofte, Radu and Li, Liang and Li, Tao and Cui, Junhui and Wang, Yunqiu and Tai, Yang and Sun, Jingwei and Sun, Jianhui and Yue, Xinli and Wang, Tianyi and Hou, Huan and Lu, Junda and Huang, Xinyang and Zhou, Zitang and Zhang, Zijian and Zheng, Xuhui and Wu, Xuecheng and Peng, Chong and Cao, Xuezhi and Nguyen-Mau, Trong-Hieu and Le, Minh-Hoang and Le-Phan, Minh-Khoa and Ly, Duy-Nam and Nguyen, Hai-Dang and Tran, Minh-Triet and Lin, Yukang and Hong, Yan and Song, Chuanbiao and Li, Siyuan and Lan, Jun and Zhang, Zhichao and Li, Xinyue and Sun, Wei and Zhang, Zicheng and Li, Yunhao and Liu, Xiaohong and Zhai, Guangtao and Xu, Zitong and Duan, Huiyu and Wang, Jiarui and Ma, Guangji and Yang, Liu and Liu, Lu and Hu, Qiang and Min, Xiongkuo and Wang, Zichuan and Tang, Zhenchen and Peng, Bo and Dong, Jing and Guan, Fengbin and Yu, Zihao and Lu, Yiting and Luo, Wei and Li, Xin and Lin, Minhao and Chen, Haofeng and He, Xuanxuan and Xu, Kele and Xu, Qisheng and Gao, Zijian and Wan, Tianjiao and Qiu, Bo-Cheng and Hsu, Chih-Chung and Lee, Chia-ming and Lin, Yu-Fan and Yu, Bo and Wang, Zehao and Mu, Da and Chen, Mingxiu and Fang, Junkang and Sun, Huamei and Zhao, Wending and Wang, Zhiyu and Liu, Wang and Yu, Weikang and Duan, Puhong and Sun, Bin and Kang, Xudong and Li, Shutao and He, Shuai and Fu, Lingzhi and Cong, Heng and Zhang, Rongyu and He, Jiarong and Qiao, Zhishan and Huang, Yongqing and Chen, Zewen and Pang, Zhe and Wang, Juan and Guo, Jian and Shao, Zhizhuo and Feng, Ziyu and Li, Bing and Hu, Weiming and Li, Hesong and Liu, Dehua and Liu, Zeming and Xie, Qingsong and Wang, Ruichen and Li, Zhihao and Liang, Yuqi and Bi, Jianqi and Luo, Jun and Yang, Junfeng and Li, Can and Fu, Jing and Xu, Hongwei and Long, Mingrui and Tang, Lulin}, title = {NTIRE 2025 challenge on Text to Image Generation Model Quality Assessment}, booktitle = {Proceedings of the Computer Vision and Pattern Recognition Conference (CVPR) Workshops}, month = {June}, year = {2025}, pages = {1104-1125} }
NTIRE 2025 challenge on Text to Image Generation Model Quality Assessment
Abstract
This paper reports on the NTIRE 2025 challenge on Text to Image (T2I) generation model quality assessment, which will be held in conjunction with the New Trends in Image Restoration and Enhancement Workshop (NTIRE) at CVPR 2025. The aim of this challenge is to address the fine-grained quality assessment of text-to-image generation models. This challenge evaluates text-to-image models from two aspects: image-text alignment and image structural distortion detection, and is divided into the alignment track and the structural track. The alignment track uses the EvalMuse-40K, which contains around 40K AI-Generated Images (AIGIs) generated by 20 popular generative models. The alignment track has a total of 371 registered participants. A total of 1,883 submissions are received in the development phase, and 507 submissions are received in the test phase. Finally, 12 participating teams submitted their models and fact sheets. The structure track uses the EvalMuse-Structure, which contains 10,000 AI-Generated Images (AIGIs) with corresponding structural distortion mask. A total of 211 participants have registered in the structure track. A total of 1155 submissions are received in the development phase, and 487 submissions are received in the test phase. Finally, 8 participating teams submitted their models and fact sheets. Almost all methods have achieved better results than baseline methods, and the winning methods in both tracks have demonstrated superior prediction performance on T2I model quality assessment.
Related Material