{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/self-adaptive-single-and-multi-illuminant","title":"Self-adaptive Single and Multi-illuminant Estimation Framework based on Deep Learning","arxiv_id":"1902.04705","date":"2019-02-13","proceeding":null,"authors":["Yongjie Liu","Sijie Shen"],"abstract":"Illuminant estimation plays a key role in digital camera pipeline system, it\naims at reducing color casting effect due to the influence of non-white\nilluminant. Recent researches handle this task by using Convolution Neural\nNetwork (CNN) as a mapping function from input image to a single illumination\nvector. However, global mapping approaches are difficult to deal with scenes\nunder multi-light-sources. In this paper, we proposed a self-adaptive single\nand multi-illuminant estimation framework, which includes the following\nnovelties: (1) Learning local self-adaptive kernels from the entire image for\nilluminant estimation with encoder-decoder CNN structure; (2) Providing\nconfidence measurement for the prediction; (3) Clustering-based iterative\nfitting for computing single and multi-illumination vectors. The proposed\nglobal-to-local aggregation is able to predict multi-illuminant regionally by\nutilizing global information instead of training in patches, as well as brings\nsignificant improvement for single illuminant estimation. We outperform the\nstate-of-the-art methods on standard benchmarks with the largest relative\nimprovement of 16%. In addition, we collect a dataset contains over 13k images\nfor illuminant estimation and evaluation. The code and dataset is available on\nhttps://github.com/LiamLYJ/KPF_WB","url_abs":"http://arxiv.org/abs/1902.04705v1","url_pdf":"http://arxiv.org/pdf/1902.04705v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"self-adaptive-single-and-multi-illuminant","repo_url":"https://github.com/LiamLYJ/KPF_WB","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"clustering","task_name":"Clustering"},{"task_slug":"decoder","task_name":"Decoder"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1902.04705","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}