ICCV 2017 Tutorial

Covariance 2017

Venice - Sunday October 22

Schedule

  • 8:30 Data representation by covariance matrices
  • 9:00 Geometry of SPD matrices
    • Euclidean distance
    • Affine-invariant Riemannian metric
    • Log-Euclidean metric
    • Bregman divergences
  • 9:45 Applications of covariance matrices in computer vision

  • Slides for Part I: Covariance Matrices and Applications

  • 10:15 Morning Break

  • 11:00 Data representation by covariance operators
    • Positive definite kernels and feature maps
    • Covariance operators of feature maps
  • 11:30 Geometry of covariance operators
    • Hilbert-Schmidt distance
    • Affine-invariant Riemannian distance
    • Log-Hilbert-Schmidt distance
    • Bregman divergences
  • 12:15 Applications of covariance operators in computer vision
    • Kernel methods with covariance operators
    • Two-layer kernel machines with covariance operators
    • Comparison with covariance matrices (performance and computational cost)
    Slides for Part II: Covariance Operators and Applications

  • Conclusion and future outlook
IIT v4 logo pavis t1