BRDF Estimation of Complex Materials with Nested Learning

The estimation of the optical properties of a material from RGB-images is an important but extremely ill-posed problem in Computer Graphics. While recent works have successfully approached this problem even from just a single photograph, significant simplifications of the material model are assumed, limiting the usability of such methods. The detection of complex material properties such as anisotropy or Fresnel effect remains an unsolved challenge. We propose a novel method that predicts the model parameters of an artist-friendly, physically-based BRDF, from only two low-resolution shots of the material. Thanks to a novel combination of deep neural networks in a nested architecture, we are able to handle the ambiguities given by the non-orthogonality and non-convexity of the parameter space. To train the network, we generate a novel dataset of physically-based synthetic images. We prove that our model can recover new properties like anisotropy, index of refraction and a second reflectance color, for materials that have tinted specular reflections or whose albedo changes at glancing angles.

[1]  Jaakko Lehtinen,et al.  Reflectance modeling by neural texture synthesis , 2016, ACM Trans. Graph..

[2]  Rich Caruana,et al.  Multitask Learning , 1998, Encyclopedia of Machine Learning and Data Mining.

[3]  Donald P. Greenberg,et al.  Toward a psychophysically-based light reflection model for image synthesis , 2000, SIGGRAPH.

[4]  Matthias Nießner,et al.  A Lightweight Approach for On-the-Fly Reflectance Estimation , 2017, 2017 IEEE International Conference on Computer Vision (ICCV).

[5]  Jirí Filip,et al.  Effective Acquisition of Dense Anisotropic BRDF , 2014, 2014 22nd International Conference on Pattern Recognition.

[6]  Jaakko Lehtinen,et al.  Practical SVBRDF capture in the frequency domain , 2013, ACM Trans. Graph..

[7]  Jannik Boll Nielsen,et al.  On optimal, minimal BRDF sampling for reflectance acquisition , 2015, ACM Trans. Graph..

[8]  Mario Fritz,et al.  Reflectance and Natural Illumination from Single-Material Specular Objects Using Deep Learning , 2018, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[9]  James F. Blinn,et al.  Models of light reflection for computer synthesized pictures , 1977, SIGGRAPH.

[10]  B K Horn,et al.  Calculating the reflectance map. , 1979, Applied optics.

[11]  Pierre Poulin,et al.  Extracting Microfacet‐based BRDF Parameters from Arbitrary Materials with Power Iterations , 2015, Comput. Graph. Forum.

[12]  Kun Zhou,et al.  Simultaneous Localization and Appearance Estimation with a Consumer RGB-D Camera , 2016, IEEE Transactions on Visualization and Computer Graphics.

[13]  Jirí Filip,et al.  Template‐Based Sampling of Anisotropic BRDFs , 2014, Comput. Graph. Forum.

[14]  Noah Snavely,et al.  OpenSurfaces , 2013, ACM Trans. Graph..

[15]  Peter Shirley,et al.  An Anisotropic Phong BRDF Model , 2000, J. Graphics, GPU, & Game Tools.

[16]  Jannik Boll Nielsen,et al.  Minimal BRDF sampling for two-shot near-field reflectance acquisition , 2016, ACM Trans. Graph..

[17]  Wojciech Matusik,et al.  Inverse shade trees for non-parametric material representation and editing , 2006, ACM Trans. Graph..

[18]  Bui Tuong Phong Illumination for computer generated pictures , 1975, Commun. ACM.

[19]  Jaakko Lehtinen,et al.  Two-shot SVBRDF capture for stationary materials , 2015, ACM Trans. Graph..

[20]  Jian Wang,et al.  Reflectance Capture Using Univariate Sampling of BRDFs , 2017, 2017 IEEE International Conference on Computer Vision (ICCV).

[21]  Eric P. Lafortune,et al.  Using the modified Phong reflectance model for physically based rendering , 1994 .

[22]  Gabriel Peyré,et al.  Wasserstein barycentric coordinates , 2016, ACM Trans. Graph..

[23]  Vlastimil Havran,et al.  BRDF Slices: Accurate Adaptive Anisotropic Appearance Acquisition , 2013, 2013 IEEE Conference on Computer Vision and Pattern Recognition.

[24]  Ko Nishino,et al.  Reflectance and Natural Illumination from a Single Image , 2012, ECCV.

[25]  Gregory J. Ward,et al.  Measuring and modeling anisotropic reflection , 1992, SIGGRAPH.

[26]  Giuseppe Claudio Guarnera,et al.  BRDF Representation and Acquisition , 2016, Comput. Graph. Forum.

[27]  Ye Yu,et al.  PVNN: A Neural Network Library for Photometric Vision , 2017, 2017 IEEE International Conference on Computer Vision Workshops (ICCVW).

[28]  Ko Nishino,et al.  Radiometric Scene Decomposition: Scene Reflectance, Illumination, and Geometry from RGB-D Images , 2016, 2016 Fourth International Conference on 3D Vision (3DV).

[29]  Tim Weyrich,et al.  Decomposing Single Images for Layered Photo Retouching , 2017, Comput. Graph. Forum.

[30]  P. Cochat,et al.  Et al , 2008, Archives de pediatrie : organe officiel de la Societe francaise de pediatrie.

[31]  Pieter Peers,et al.  genBRDF: discovering new analytic BRDFs with genetic programming , 2014, ACM Trans. Graph..

[32]  Alexander Wilkie,et al.  Arbitrarily layered micro-facet surfaces , 2007, GRAPHITE '07.

[33]  Ersin Yumer,et al.  Material Editing Using a Physically Based Rendering Network , 2017, 2017 IEEE International Conference on Computer Vision (ICCV).

[34]  Frédo Durand,et al.  Experimental analysis of BRDF models , 2005, EGSR '05.

[35]  Jirí Filip,et al.  Minimal Sampling for Effective Acquisition of Anisotropic BRDFs , 2016, Comput. Graph. Forum.

[36]  Xiao Li,et al.  Modeling surface appearance from a single photograph using self-augmented convolutional neural networks , 2017, ACM Trans. Graph..

[37]  Hans-Peter Seidel,et al.  LIME: Live Intrinsic Material Estimation , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.