Flexible SVBRDF Capture with a Multi-Image Deep Network

Abstract : Empowered by deep learning, recent methods for material capture can estimate a spatially-varying reflectance from a single photograph. Such lightweight capture is in stark contrast with the tens or hundreds of pictures required by traditional optimization-based approaches. However, a single image is often simply not enough to observe the rich appearance of real-world materials. We present a deep-learning method capable of estimating material appearance from a variable number of uncalibrated and unordered pictures captured with a handheld camera and flash. Thanks to an order-independent fusing layer, this architecture extracts the most useful information from each picture, while benefiting from strong priors learned from data. The method can handle both view and light direction variation without calibration. We show how our method improves its prediction with the number of input pictures, and reaches high quality reconstructions with as little as 1 to 10 images-a sweet spot between existing single-image and complex multi-image approaches.
Document type :
Journal articles
Complete list of metadatas

Cited literature [40 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-02164993
Contributor : Team Reves <>
Submitted on : Thursday, June 27, 2019 - 1:10:27 PM
Last modification on : Wednesday, October 23, 2019 - 5:00:50 PM

File

multi_inputs_EGSR.pdf
Files produced by the author(s)

Identifiers

Collections

Citation

Valentin Deschaintre, Miika Aittala, Frédo Durand, George Drettakis, Adrien Bousseau. Flexible SVBRDF Capture with a Multi-Image Deep Network. Computer Graphics Forum, Wiley, 2019, 38, ⟨10.1111/cgf.13765⟩. ⟨hal-02164993v2⟩

Share

Metrics

Record views

97

Files downloads

122