Learned Smartphone Isp on Mobile Gpus, Mobile Ai 2025 Challenge: Report
Andrey Ignatov, Georgy Perevozchikov, Radu Timofte, Li Cheng, Lian Liu, Jun Cao, Heng Sun, Pan Wu, Song Wang, K. W. Yu, Shuo Liu, Haiping He, ZhenHao Dong, JianKe Chen, Dejun Hao, Keqiang Yu, Tingniao Wang, Xiaoqing Zhou, Dong Zhang, Chunxia Lei, Jianke Chen, Dejun Hao, J. He, Hailong Yan, Ao Li, Xiangtao Zhang, Zhe Liu, Ce Zhu, Le Zhang, Andrei Arhire, Shuo Liu, Junpyo Seo, Zhenhao Dong, Jianke Chen, Fen Xie, Xiuzhi Fang, Chen Wu, Zhansheng Wang, Pengbo Zhang, Jiazi Huang
Abstract
RGB photo reconstruction from RAW camera images is an increasingly popular deep learning problem with a practical application to mobile cameras. This creates a need for solutions that are not only performant but are additionally compatible with real mobile AI hardware such as GPUs or NPUs. In this Mobile AI challenge, we address this problem and propose the participants to design efficient learned ISP models that can demonstrate fast inference times on mobile GPUs. For this, the participants were provided with a largescale Fujifilm UltraISP dataset consisting of RAW-RGB image pairs captured with the Sony IMX586 Quad Bayer mobile sensor and a professional 102MP medium format FujiFilm GFX100 camera. The runtime of all models was evaluated on the latest Adreno and Mali GPUs used in Qualcomm and MediaTek chipsets. The proposed solutions are compatible with all recent mobile GPUs, being able to process Full HD photos under 30 ms in the majority of cases and delivering high-fidelity results. A comprehensive description of the models developed in the challenge is provided in this paper.