DeepBRDF: A Deep Representation for Manipulating Measured BRDF

Simple item page
dc.contributor.authorHu, Bingyangen_US
dc.contributor.authorGuo, Jieen_US
dc.contributor.authorChen, Yanjunen_US
dc.contributor.authorLi, Mengtianen_US
dc.contributor.authorGuo, Yanwenen_US
dc.contributor.editorPanozzo, Daniele and Assarsson, Ulfen_US
dc.date.accessioned2020-05-24T12:51:22Z
dc.date.available2020-05-24T12:51:22Z
dc.date.issued2020
dc.description.abstractEffective compression of densely sampled BRDF measurements is critical for many graphical or vision applications. In this paper, we present DeepBRDF, a deep-learning-based representation that can significantly reduce the dimensionality of measured BRDFs while enjoying high quality of recovery. We consider each measured BRDF as a sequence of image slices and design a deep autoencoder with a masked L2 loss to discover a nonlinear low-dimensional latent space of the high-dimensional input data. Thorough experiments verify that the proposed method clearly outperforms PCA-based strategies in BRDF data compression and is more robust. We demonstrate the effectiveness of DeepBRDF with two applications. For BRDF editing, we can easily create a new BRDF by navigating on the low-dimensional manifold of DeepBRDF, guaranteeing smooth transitions and high physical plausibility. For BRDF recovery, we design another deep neural network to automatically generate the full BRDF data from a single input image. Aided by our DeepBRDF learned from real-world materials, a wide range of reflectance behaviors can be recovered with high accuracy.en_US
dc.description.number2
dc.description.sectionheadersDeep Learning for Rendering
dc.description.seriesinformationComputer Graphics Forum
dc.description.volume39
dc.identifier.doi10.1111/cgf.13920
dc.identifier.issn1467-8659
dc.identifier.pages157-166
dc.identifier.urihttps://doi.org/10.1111/cgf.13920
dc.identifier.urihttps://diglib.eg.org:443/handle/10.1111/cgf13920
dc.publisherThe Eurographics Association and John Wiley & Sons Ltd.en_US
dc.rightsAttribution 4.0 International License
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/
dc.subjectComputing methodologies
dc.subjectReflectance modeling
dc.subjectNeural networks
dc.titleDeepBRDF: A Deep Representation for Manipulating Measured BRDFen_US
Files
Original bundle
Now showing 1 - 2 of 2
Thumbnail Image
Name:
v39i2pp157-166.pdf
Size:
28.79 MB
Format:
Adobe Portable Document Format
Download
Thumbnail Image
Name:
supp.pdf
Size:
40.54 MB
Format:
Adobe Portable Document Format
Download
Collections
39-Issue 2