SVBRDF Recovery from a Single Image with Highlights Using a Pre‐trained Generative Adversarial Network - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Journal Articles Computer Graphics Forum Year : 2022

SVBRDF Recovery from a Single Image with Highlights Using a Pre‐trained Generative Adversarial Network

Abstract

Input SVBRDF maps SVBRDF maps Render Render Figure 1: Our method generates high-quality and natural SVBRDF maps from a single input photograph with overexposed highlight regions, and provides vivid rendering. Our Fourier-based loss function separates the effects due to the specular highlight from those due to the material.
Fichier principal
Vignette du fichier
paper_main.pdf (36.87 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03937939 , version 1 (13-01-2023)

Licence

Attribution

Identifiers

Cite

Tao Wen, Beibei Wang, Lei Zhang, Jie Guo, Nicolas Holzschuch. SVBRDF Recovery from a Single Image with Highlights Using a Pre‐trained Generative Adversarial Network. Computer Graphics Forum, 2022, 41 (6), pp.110-123. ⟨10.1111/cgf.14514⟩. ⟨hal-03937939⟩
41 View
8 Download

Altmetric

Share

Gmail Facebook X LinkedIn More