Loading...
Data-Driven Reflectance Estimation Under Natural Lighting
Cooper, Victoria
Cooper, Victoria
Abstract
Bidirectional Reflectance Distribution Functions, (BRDFs), describe how light is reflected off of a material. BRDFs are captured so that the materials can be re-lit under new while maintaining accuracy. BRDF models can approximate the reflectance of a material, but are unable to accurately represent the full BRDF of the material. Acquisition setups for BRDFs trade accuracy for speed with the most accurate methods, gonioreflectometers, being the slowest. Image-based BRDF acquisition approaches range from using complicated controlled lighting setups to uncontrolled known lighting to assuming the lighting is unknown. We propose a data-driven method for recovering BRDFs under known, but uncontrolled lighting. This approach utilizes a dataset of 100 measured BRDFs to accurately reconstruct the BRDF from a single photograph. We model the BRDFs as Gaussian Mixture Models, (GMMs), and use an Expectation Maximization, (EM), approach to determine cluster membership. We apply this approach to captured data as well as synthetic. We continue this work by relaxing assumptions about either lighting, material, or geometry. This work was supported in part by NSF grant IIS-1350323 and gifts from Google, Activision, and Nvidia.
Description
Date
2021-07-01
Journal Title
Journal ISSN
Volume Title
Publisher
Collections
Download Dataset
Files
Rights Holder
Usage License
Embargo
Research Projects
Organizational Units
Journal Issue
Keywords
Citation
Department
Computer Science
DOI
http://dx.doi.org/10.21220/s2-fwpk-zr81