Learned Smartphone ISP on Mobile GPUs with Deep Learning, Mobile AI & AIM 2022 Challenge: Report
- Authors
- Ignatov, A.; Timofte, R.; Liu, S.; Feng, C.; Bai, F.; Wang, X.; Lei, L.; Yi, Z.; Xiang, Y.; Liu, Z.; Li, S.; Shi, K.; Kong, D.; Xu, K.; Kwon, M.; Wu, Y.; Zheng, J.; Fan, Z.; Wu, X.; Zhang, F.; No, A.; Cho, M.; Chen, Z.; Zhang, X.; Li, R.; Wang, J.; Wang, Z.; Conde, M.V.; Choi, U.-J.; Perevozchikov, G.; Ershov, E.; Hui, Z.; Dong, M.; Lou, X.; Zhou, W.; Pang, C.; Qin, H.; Cai, M.
- Issue Date
- 1-Jan-2023
- Publisher
- Springer Science and Business Media Deutschland GmbH
- Keywords
- AI Benchmark; Deep learning; Learned ISP; Mobile AI; Mobile AI Challenge; Mobile cameras; Photo enhancement
- Citation
- Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), v.13803 LNCS, pp.44 - 70
- Journal Title
- Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
- Volume
- 13803 LNCS
- Start Page
- 44
- End Page
- 70
- URI
- https://scholarworks.bwise.kr/hongik/handle/2020.sw.hongik/31066
- DOI
- 10.1007/978-3-031-25066-8_3
- ISSN
- 0302-9743
- Abstract
- The role of mobile cameras increased dramatically over the past few years, leading to more and more research in automatic image quality enhancement and RAW photo processing. In this Mobile AI challenge, the target was to develop an efficient end-to-end AI-based image signal processing (ISP) pipeline replacing the standard mobile ISPs that can run on modern smartphone GPUs using TensorFlow Lite. The participants were provided with a large-scale Fujifilm UltraISP dataset consisting of thousands of paired photos captured with a normal mobile camera sensor and a professional 102MP medium-format FujiFilm GFX100 camera. The runtime of the resulting models was evaluated on the Snapdragon’s 8 Gen 1 GPU that provides excellent acceleration results for the majority of common deep learning ops. The proposed solutions are compatible with all recent mobile GPUs, being able to process Full HD photos in less than 20–50 ms while achieving high fidelity results. A detailed description of all models developed in this challenge is provided in this paper. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - College of Engineering > School of Electronic & Electrical Engineering > 1. Journal Articles
![qrcode](https://api.qrserver.com/v1/create-qr-code/?size=55x55&data=https://scholarworks.bwise.kr/hongik/handle/2020.sw.hongik/31066)
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.