De-lighting a High-resolution Picture for Material Acquisition

Simple item page
dc.contributor.authorMartin, Rosalieen_US
dc.contributor.authorMeyer, Arthuren_US
dc.contributor.authorPesare, Davideen_US
dc.contributor.editorBoubekeur, Tamy and Sen, Pradeepen_US
dc.date.accessioned2019-07-14T19:22:48Z
dc.date.available2019-07-14T19:22:48Z
dc.date.issued2019
dc.description.abstractWe propose a deep-learning based method for the removal of shades, projected shadows and highlights from a single picture of a quasi-planar surface captured in natural lighting conditions with any kind of camera device. To achieve this, we train an encoder-decoder to process physically based materials, rendered under various lighting conditions, to infer its spatially-varying albedo. Our network processes relatively small image tiles (512x512 pixels) and we propose a solution to handle larger image resolutions by solving a Poisson system across these tiles.en_US
dc.description.sectionheadersIndustry Track
dc.description.seriesinformationEurographics Symposium on Rendering - DL-only and Industry Track
dc.identifier.doi10.2312/sr.20191222
dc.identifier.isbn978-3-03868-095-6
dc.identifier.issn1727-3463
dc.identifier.pages69-72
dc.identifier.urihttps://doi.org/10.2312/sr.20191222
dc.identifier.urihttps://diglib.eg.org:443/handle/10.2312/sr20191222
dc.publisherThe Eurographics Associationen_US
dc.titleDe-lighting a High-resolution Picture for Material Acquisitionen_US
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
069-072.pdf
Size:
9.95 MB
Format:
Adobe Portable Document Format
Description:
Download
Collections
Rendering 2019 - DL-only / Industry Track