convex concave convex concave Eigenfaces Photobook/Eigenfaces (MIT Media Lab)

  • View
    213

  • Download
    0

Embed Size (px)

Text of convex concave convex concave Eigenfaces Photobook/Eigenfaces (MIT Media Lab)

  • Slide 1
  • Slide 2
  • Slide 3
  • Slide 4
  • convex concave
  • Slide 5
  • convex concave
  • Slide 6
  • Slide 7
  • Slide 8
  • Eigenfaces Photobook/Eigenfaces (MIT Media Lab)
  • Slide 9
  • Database 7562 pictures of 3000 people Photobook/Eigenfaces (MIT Media Lab)
  • Slide 10
  • Query Example Photobook/Eigenfaces (MIT Media Lab)
  • Slide 11
  • Eigenfeatures Photobook/Eigenfaces (MIT Media Lab)
  • Slide 12
  • Eigenfeatures
  • Slide 13
  • Photobook/Eigenfaces (MIT Media Lab) Eigenfeatures
  • Slide 14
  • Receiver Operating Characteristic (ROC) Curve Photobook/Eigenfaces (MIT Media Lab) Eigenfeatures
  • Slide 15
  • Recognition with PCA Amano, Hiura, Yamaguti, and Inokuchi; Atick and Redlich; Bakry, Abo-Elsoud, and Kamel; Belhumeur, Hespanha, and Kriegman; Bhatnagar, Shaw, and Williams; Black and Jepson; Brennan and Principe; Campbell and Flynn; Casasent, Sipe and Talukder; Chan, Nasrabadi and Torrieri; Chung, Kee and Kim; Cootes, Taylor, Cooper and Graham; Covell; Cui and Weng; Daily and Cottrell; Demir, Akarun, and Alpaydin; Duta, Jain and Dubuisson-Jolly; Hallinan; Han and Tewfik; Jebara and Pentland; Kagesawa, Ueno, Kasushi, and Kashiwagi; King and Xu; Kalocsai, Zhao, and Elagin; Lee, Jung, Kwon and Hong; Liu and Wechsler; Menser and Muller; Moghaddam; Moon and Philips; Murase and Nayar; Nishino, Sato, and Ikeuchi; Novak, and Owirka; Nishino, Sato, and Ikeuchi; Ohta, Kohtaro and Ikeuchi; Ong and Gong; Penev and Atick; Penev and Sirivitch; Lorente and Torres; Pentland, Moghaddam, and Starner; Ramanathan, Sum, and Soon; Reiter and Matas; Romdhani, Gong and Psarrou; Shan, Gao, Chen, and Ma; Shen, Fu, Xu, Hsu, Chang, and Meng; Sirivitch and Kirby; Song, Chang, and Shaowei; Torres, Reutter, and Lorente; Turk and Pentland; Watta, Gandhi, and Lakshmanan; Weng and Chen; Yuela, Dai, and Feng; Yuille, Snow, Epstein, and Belhumeur; Zhao, Chellappa, and Krishnaswamy; Zhao and Yang.
  • Slide 16
  • Lambertian Reflectance Matt surface Light source is distant Light reflected equally to all directions or
  • Slide 17
  • Photometric Stereo: Factorization M is f x p (#images x #pixels) L is f x 3 light sources S is 3 x p surface normals (scaled by albedo) Rank(M)=3 (if no noise present) SVD: Ambiguity Eliminate by forcing integrability
  • Slide 18
  • Relief Sculptures
  • Slide 19
  • Illumination Cone =0.5*+0.2*+0.3*
  • Slide 20
  • Empirical Study BallFacePhoneParrot #148.253.767.942.8 #394.490.288.276.3 #597.993.594.184.7 #799.195.396.388.5 #999.596.397.290.7 (Yuille et al.) Dimension:
  • Slide 21
  • BallFacePhoneParrot #148.253.767.942.8 #284.475.283.269.7 #394.490.288.276.3 #496.592.192.081.5 #597.993.594.184.7 #698.994.595.287.2 #799.195.396.388.5 #899.395.896.889.7 #999.596.397.290.7 #1099.696.697.591.7
  • Slide 22
  • Intuition lighting reflectance
  • Slide 23
  • Spherical Harmonics Orthonormal basis for functions on the sphere nth order harmonics have 2n+1 components Rotation = phase shift (same n, different m) In space coordinates: polynomials of degree n Funk-Hecke convolution theorem
  • Slide 24
  • Spherical Harmonics ZYX XZYZXY
  • Slide 25
  • Harmonic Transform of Kernel n
  • Slide 26
  • Cumulative Energy N (percents)
  • Slide 27
  • Second Order Approximation
  • Slide 28
  • Other Low-D Approximations HemisphereForeshortenedBall (Exp.)Face ModelFace (Exp.) #15162486154 #26977848275 #38892949290 #49395979692 #59597989794 #698 999895 #79899 95 #899 96 #999 1009996 (Ramamoorthi)
  • Slide 29
  • Harmonic Images
  • Slide 30
  • Reconstruction
  • Slide 31
  • Slide 32
  • Motion + Illumination
  • Slide 33
  • Reconstruction Reconstruction Laser scan
  • Slide 34
  • Advantage of Our Method Disparity error Residue Std intensity Accounting for illumination variation Assuming brightness constancy
  • Slide 35
  • Mutual Information (Viola and Wells) Camera Rotation