1 Surface Reflectance Estimation and Natural Illumination Statistics Ron Dror, Ted Adelson, Alan Willsky Artificial Intelligence Lab, Lab for Information.

Slides:



Advertisements
Similar presentations
Computer Vision Radiometry. Bahadir K. Gunturk2 Radiometry Radiometry is the part of image formation concerned with the relation among the amounts of.
Advertisements

Evaluating Color Descriptors for Object and Scene Recognition Koen E.A. van de Sande, Student Member, IEEE, Theo Gevers, Member, IEEE, and Cees G.M. Snoek,
1 Graphics CSCI 343, Fall 2013 Lecture 18 Lighting and Shading.
Environment Mapping CSE 781 – Roger Crawfis
Spherical Convolution in Computer Graphics and Vision Ravi Ramamoorthi Columbia Vision and Graphics Center Columbia University SIAM Imaging Science Conference:
Capturing light Source: A. Efros. Image formation How bright is the image of a scene point?
Computer Vision CS 776 Spring 2014 Photometric Stereo and Illumination Cones Prof. Alex Berg (Slide credits to many folks on individual slides)
Measuring BRDFs. Why bother modeling BRDFs? Why not directly measure BRDFs? True knowledge of surface properties Accurate models for graphics.
1 Computer Graphics Chapter 9 Rendering. [9]-2RM Rendering Three dimensional object rendering is the set of collective processes which make the object.
Foundations of Computer Graphics (Spring 2012) CS 184, Lecture 21: Radiometry Many slides courtesy Pat Hanrahan.
Basic Principles of Surface Reflectance
Acquiring the Reflectance Field of a Human Face Paul Debevec, Tim Hawkins, Chris Tchou, Haarm-Pieter Duiker, Westley Sarokin, Mark Sagar Haarm-Pieter Duiker,
Advanced Computer Graphics (Spring 2013) CS 283, Lecture 8: Illumination and Reflection Many slides courtesy.
電腦視覺 Computer and Robot Vision I Chapter2: Binary Machine Vision: Thresholding and Segmentation Instructor: Shih-Shinh Huang 1.
1. What is Lighting? 2 Example 1. Find the cubic polynomial or that passes through the four points and satisfies 1.As a photon Metal Insulator.
December 5, 2013Computer Vision Lecture 20: Hidden Markov Models/Depth 1 Stereo Vision Due to the limited resolution of images, increasing the baseline.
A Signal-Processing Framework for Inverse Rendering Ravi Ramamoorthi Pat Hanrahan Stanford University.
A Signal-Processing Framework for Forward and Inverse Rendering COMS , Lecture 8.
An Efficient Representation for Irradiance Environment Maps Ravi Ramamoorthi Pat Hanrahan Stanford University.
Recovering Intrinsic Images from a Single Image 28/12/05 Dagan Aviv Shadows Removal Seminar.
Computational Fundamentals of Reflection COMS , Lecture
Computer Graphics (Spring 2008) COMS 4160, Lecture 20: Illumination and Shading 2
A Signal-Processing Framework for Forward and Inverse Rendering Ravi Ramamoorthi Ravi Ramamoorthi
Measurement, Inverse Rendering COMS , Lecture 4.
A Signal-Processing Framework for Forward and Inverse Rendering Ravi Ramamoorthi Ph.D. Orals: June 3, 2002.
Face Recognition Based on 3D Shape Estimation
Computer Graphics (Fall 2008) COMS 4160, Lecture 19: Illumination and Shading 2
Basic Principles of Surface Reflectance
Rendering General BSDFs and BSSDFs Randy Rauwendaal.
Computational Vision Jitendra Malik University of California at Berkeley Jitendra Malik University of California at Berkeley.
A Signal-Processing Framework for Forward and Inverse Rendering Ravi Ramamoorthi Stanford University Columbia University: Feb 11, 2002.
Basic Principles of Surface Reflectance Lecture #3 Thanks to Shree Nayar, Ravi Ramamoorthi, Pat Hanrahan.
Computer Graphics (Fall 2004) COMS 4160, Lecture 16: Illumination and Shading 2 Lecture includes number of slides from.
Object recognition under varying illumination. Lighting changes objects appearance.
Aydın ÖZTÜRK Murat KURT Ahmet BİLGİLİ GraphiCon’ th International Conference on Computer Graphics and Vision September 20-24, 2010, St.Petersburg,
Measure, measure, measure: BRDF, BTF, Light Fields Lecture #6
Specular Reflections from Rough Surfaces Lecture #4 Thanks to Shree Nayar, Ravi Ramamoorthi, Pat Hanrahan.
Computer Graphics Inf4/MSc Computer Graphics Lecture Notes #16 Image-Based Lighting.
Light and shading Source: A. Efros.
CS 445 / 645: Introductory Computer Graphics
Shading / Light Thanks to Srinivas Narasimhan, Langer-Zucker, Henrik Wann Jensen, Ravi Ramamoorthi, Hanrahan, Preetham.
Image-Based Rendering from a Single Image Kim Sang Hoon Samuel Boivin – Andre Gagalowicz INRIA.
-Global Illumination Techniques
Texture. Texture is an innate property of all surfaces (clouds, trees, bricks, hair etc…). It refers to visual patterns of homogeneity and does not result.
Real-Time Rendering Digital Image Synthesis Yung-Yu Chuang 01/03/2006 with slides by Ravi Ramamoorthi and Robin Green.
December 4, 2014Computer Vision Lecture 22: Depth 1 Stereo Vision Comparing the similar triangles PMC l and p l LC l, we get: Similarly, for PNC r and.
Capturing light Source: A. Efros.
Real-Time Relighting Digital Image Synthesis Yung-Yu Chuang 1/10/2008 with slides by Ravi Ramamoorthi, Robin Green and Milos Hasan.
Taku KomuraComputer Graphics Local Illumination and Shading Computer Graphics – Lecture 10 Taku Komura Institute for Perception, Action.
Inverse Global Illumination: Recovering Reflectance Models of Real Scenes from Photographs Computer Science Division University of California at Berkeley.
Face Detection Using Large Margin Classifiers Ming-Hsuan Yang Dan Roth Narendra Ahuja Presented by Kiang “Sean” Zhou Beckman Institute University of Illinois.
Interreflections : The Inverse Problem Lecture #12 Thanks to Shree Nayar, Seitz et al, Levoy et al, David Kriegman.
Announcements Office hours today 2:30-3:30 Graded midterms will be returned at the end of the class.
Photo-realistic Rendering and Global Illumination in Computer Graphics Spring 2012 Material Representation K. H. Ko School of Mechatronics Gwangju Institute.
Chapter1: Introduction Chapter2: Overview of Supervised Learning
Computer Graphics (Spring 2003) COMS 4160, Lecture 18: Shading 2 Ravi Ramamoorthi Guest Lecturer: Aner Benartzi.
Cornell CS465 Spring 2004 Lecture 4© 2004 Steve Marschner 1 Shading CS 465 Lecture 4.
1Ellen L. Walker 3D Vision Why? The world is 3D Not all useful information is readily available in 2D Why so hard? “Inverse problem”: one image = many.
Illumination Model How to compute color to represent a scene As in taking a photo in real life: – Camera – Lighting – Object Geometry Material Illumination.
Announcements Final is Thursday, March 18, 10:30-12:20 –MGH 287 Sample final out today.
EECS 274 Computer Vision Sources, Shadows, and Shading.
MAN-522 Computer Vision Spring
Shading CS 465 Lecture 4 © 2004 Steve Marschner • 1.
Computer Graphics Chapter 9 Rendering.
Real-time environment map lighting
Shading Variation in observed color across an object
Common Classification Tasks
Image Based Modeling and Rendering (PI: Malik)
Filtering An image as a function Digital vs. continuous images
Shape from Shading and Texture
Presentation transcript:

1 Surface Reflectance Estimation and Natural Illumination Statistics Ron Dror, Ted Adelson, Alan Willsky Artificial Intelligence Lab, Lab for Information and Decision Systems July 13, 2001

2 Reflectance Estimation Problem Surface appearance depends on surface reflectance, illumination, and geometry. We wish to estimate reflectance under unknown illumination.

3 Human vision

4 Machine vision

5 Motivation Recognize materials. Reflectance, like texture, is a primary visual characteristic of materials. Material recognition is important in its own right and as a complement to shape recognition. Capture real-world reflectances for rendering purposes. Rectify classical motion, stereo, and shape- from-shading algorithms.

6 Reflectance estimation is ill-posed A surface’s BRDF f(  i,  i ;  r,  r ) specifies how much of the light incident from any one direction is emitted in any second direction. The brightness of a surface patch to a viewer is a weighted integral over illumination from all directions. Goal: estimate reflectance (function of 4 variables) from an image (function of 2 variables) under unknown illumination from every direction (function of 2 variables at every point on the surface). More degrees of freedom than measurements, even assuming known geometry, homogeneous reflectance.

7 Bayesian formulation Find the most likely reflectance given image data. Given image data R, find most likely reflectance f by marginalizing over illumination I. P(f) – prior probability of a reflectance function P(I) – prior probability of an illumination field Challenges: P(f) and P(I) are not readily available. Integration over all illuminations is computationally daunting!

8 Two simplified formulations 1. Classification (finite but arbitrary classes): 2. Parameter estimation using a reflectance model (regression).

9 Prior information: illumination Assuming distant light sources, we can represent illumination by a single spherical image. Projection of spherical map Rendered surfaces

10 Statistical models of illumination Illumination maps possess statistical regularities akin to those of “natural images”. Histogram of pixel intensities Histogram of wavelet coefficients

11 Importance of illumination statistics for humans People recognize reflectance more easily under realistic illumination than simplified illumination.

12 Ward reflectance model A physically realizable variant of the Phong model (satisfies energy conservation and reciprocity).  d : proportion of incident radiation reflected diffusely.  s : proportion of incident radiation reflected specularly.  : surface roughness, or blur in specular component. diffuse component specular lobe

13 Effect of Ward model parameters on pixel intensity histogram Original pixel intensity probability

14 Effect of Ward model parameters on pixel intensity histogram Original  d =.1 pixel intensity probability

15 Effect of Ward model parameters on pixel intensity histogram Original  d =.2 pixel intensity probability

16 Effect of Ward model parameters on pixel intensity histogram Original  d =.3 pixel intensity probability

17 Effect of Ward model parameters on pixel intensity histogram Original  d =.4 pixel intensity probability

18 Effect of Ward model parameters on pixel intensity histogram Original  d =.4  s =.1 pixel intensity probability

19 Effect of Ward model parameters on pixel intensity histogram Original  d =.4  s =.2 pixel intensity probability

20 Effect of Ward model parameters on pixel intensity histogram Original  d =.4  s =.3 pixel intensity probability

21 Effect of Ward model parameters on pixel intensity histogram Original  d =.4  s =.4 pixel intensity probability

22 Effect of Ward model parameters on pixel intensity histogram Original  d =.4  s =.4  =0 pixel intensity probability

23 Effect of Ward model parameters on pixel intensity histogram Original  d =.4  s =.4  =.05 pixel intensity probability

24 Effect of Ward model parameters on pixel intensity histogram Original  d =.4  s =.4  =.1 pixel intensity probability

25 Effect of Ward model parameters on pixel intensity histogram Original  d =.4  s =.4  =.15 pixel intensity probability

Dependence of statistics on reflectance parameters Real-world illuminations Random checkerboard illuminations  d  s   d  s  kurt mean skew 10% 50% 90% var kurt mean skew 10% 50% 90% var kurt mean skew 10% 50% 90% var normalized derivative

27 Each reflectance clusters in feature space black matteblack shiny white mattewhite shiny gray shinychrome

28 A system for classification “Learn” relationships between features of the observed image and reflectance classes. For a distant viewer and convex object, radiance depends only on local surface orientation.

29 Implementation flow chart This leaves two open questions: How to select relevant statistics? How to build a classifier?

30 SVM classifier Support vector machines are relatively robust to the inclusion of extraneous features. A sample classifier based on just two statistics: black matteblack shiny white mattewhite shiny gray shinychrome

31 Training data sets 6 Ward model reflectances, 9 illuminations (Debevec) 11 Ward model reflectances, 100 illuminations (Teller) 9 real spheres, photographed at seven locations

32 Performance Rendered: 6 BRDFs, 9 illums Rendered: 11 BRDFs, 100 illums Photos: 9 spheres, 7 illums Chance16.7% 9.1%11.1% 6 hand- selected features 98.1%98.5%93.7% 6 auto- selected features 96.3%94.4%74.6% 6 PCA features 79.6%86.8%71.4%

33 Conclusions Our classifier rivals human performance when geometry is known and reflectance is homogeneous. Although ill-posed, reflectance estimation under unknown natural illumination is tractable. The statistical structure of natural illumination plays an essential role in visual reflectance estimation by humans and machines.

34 Future directions In progress or submitted: Extension to complex or unknown geometry; robustness to incorrect assumed geometry. Quantitative study of natural illumination statistics. Measurement of human ability to estimate reflectance from a single image without contextual information. Additional goals: Rigorous theoretical foundation – link illumination statistics directly to selected features. Estimate spatially varying reflectance.

35 Misclassifications Illumination Misclassified image Potential source of confusion

36 Feature selection By hand, based on insights developed through work with Ward model. Using automated feature selection method, which iterates the following steps: Estimate marginal probability density of each feature for each class. Select the feature that minimizes Bayes error. Regress remaining features against selected features, and subtract off predicted values.

37 Auto-selected features 6 features selected based on images of spheres with 6 Ward model reflectances under 9 illuminations: 10th percentile of 4th finest vertical subband 90th percentile of pixel intensity variance of 3rd finest diagonal subband 10th percentile of pixel intensity 90th percentile of 4th finest vertical subband median of 3rd finest horizontal subband Hand-selected features mean and 10th percentile of original image variance of two finest vertical subbands ratio of these two variances kurtosis of second finest vertical subband

38 Complex vs. simple illumination People recognize reflectance more easily under realistic illumination than simplified illumination. A reflectance estimation algorithm which takes advantage of natural illumination statistics will fail for atypical illumination.

39 Human reflectance estimation Pool balls Note ambiguity in overall color and brightness when matte spheres are viewed in isolation.

40 Related Work Yu, Debevec, Malik, and Hawkins, ’99 Reflectance and illumination from multiple photos. Sato, Wheeler, and Ikeuchi, ’97 Reflectance and geometry from photos and laser range finder, with known illumination. Marschner, Greenberg, et al., ’98, ’99 Reflectance under known illumination. Tominaga and Tanaka, 1999, ’00 Reflectance and geometry under simple lighting, using color separation. Pellacini, Ferwerda, Greenberg, ’00 Perceptually uniform gloss space for graphics. Ramamoorthi and Hanrahan, ’01 Determine when reflectance estimation problem is well- posed.

41 Photographic data Nine different spheres under the same illumination.

42 Photographic data Same spheres under a second illumination.

43 Photographic data Same spheres under a third illumination.

44 Illumination conditions

45 Rendered data set 6 spheres under one illumination condition

46 Rendered data set 6 spheres under a 2nd illumination condition

47 Is this task even possible? Humans are good at it. In psychophysical tests, we found that humans could match synthetic images of surfaces with similar reflectances rendered under different real-world illuminations. Two conclusions: Humans rely on prior information in estimating reflectance. Humans estimate reflectance without explicitly estimating illumination. Photographs of three spheres under two illumination conditions

48 Additional applications Rectify motion, stereo, and shape-from- shading algorithms. Capture real-world reflectances for rendering purposes. Yu, Debevec, Malik, Hawkins, SIGGRAPH 1999 Gideon Stein, unpublished

49 Debevec spheres

Dependence of statistics on reflectance parameters Real-world illuminations Random checkerboard illuminations  d  s   d  s  kurt mean skew 10% 50% 90% var kurt mean skew 10% 50% 90% var kurt mean skew 10% 50% 90% var normalized derivative kurt mean skew 10% 50% 90% var kurt mean skew 10% 50% 90% var kurtmeanskew 10% 50% 90% var