Image White Balance for Multi-Illuminant Scenes

Simple item page
dc.contributor.advisorKonstantinos G. Derpanis
dc.contributor.authorAditya Arora
dc.date.accessioned2024-11-07T11:15:58Z
dc.date.available2024-11-07T11:15:58Z
dc.date.copyright2024-09-13
dc.date.issued2024-11-07
dc.date.updated2024-11-07T11:15:56Z
dc.degree.disciplineComputer Science
dc.degree.levelMaster's
dc.degree.nameMSc - Master of Science
dc.description.abstractPerforming white-balance (WB) correction for scenes with multiple illuminants remains a challenging task in computer vision. Most previous methods estimate per-pixel scene illumination directly in the RAW sensor image space. Recent work explored an alternative fusion strategy, where a neural network fuses multiple white-balanced versions of the input image processed to sRGB using pre-defined white-balance settings. Inspired by this line of work, we present two contributions targeting fusion-based multi-illuminant WB correction. First, we introduce a large-scale multi-illumination dataset rendered from RAW images to support training fusion models and evaluation. The dataset comprises over 16,000 sRGB images with ground truth sRGB white-balance corrected images. Next, we introduce an attention-based architecture to fuse five white-balance settings. This architecture yields an improvement of up to 25% over prior work.
dc.identifier.urihttps://hdl.handle.net/10315/42495
dc.languageen
dc.rightsAuthor owns copyright, except where explicitly noted. Please contact the author directly with licensing requests.
dc.subjectComputer science
dc.subject.keywordsImage White-Balance
dc.subject.keywordsMulti-Illuminant Scenes
dc.subject.keywordsComputational Photography
dc.subject.keywordsComputer Vision
dc.titleImage White Balance for Multi-Illuminant Scenes
dc.typeElectronic Thesis or Dissertation

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Arora_Aditya_2024_Masters.pdf
Size:
78.43 MB
Format:
Adobe Portable Document Format
Download

Collections

ETD SWORD Deposit

All items in the YorkSpace institutional repository are protected by copyright, with all rights reserved except where explicitly noted.