Preview this item
Preview this item
Checking...

# Covariances in computer vision and machine learning

Author: Quang Minh Hà; Vittorio Murino [San Rafael, California] : Morgan & Claypool, 2018. Synthesis lectures on computer vision, # 13. eBook : Document : EnglishView all editions and formats Covariance matrices play important roles in many areas of mathematics, statistics, and machine learning, as well as their applications. In computer vision and image processing, they give rise to a powerful data representation, namely the covariance descriptor, with numerous practical applications. In this book, we begin by presenting an overview of the finite-dimensional covariance matrix representation approach of images, along with its statistical interpretation. In particular, we discuss the various distances and divergences that arise from the intrinsic geometrical structures of the set of Symmetric Positive Definite (SPD) matrices, namely Riemannian manifold and convex cone structures. Computationally, we focus on kernel methods on covariance matrices, especially using the Log-Euclidean distance. We then show some of the latest developments in the generalization of the finite dimensional covariance matrix representation to the infinite-dimensional covariance operator representation via positive definite kernels. We present the generalization of the affine-invariant Riemannian metric and the Log-Hilbert-Schmidt metric, which generalizes the Log-Euclidean distance. Computationally, we focus on kernel methods on covariance operators, especially using the Log-Hilbert-Schmidt distance. Specifically, we present a two-layer kernel machine, using the Log-Hilbert-Schmidt distance and its finite-dimensional approximation, which reduces the computational complexity of the exact formulation while largely preserving its capability. Theoretical analysis shows that, mathematically, the approximate Log-Hilbert-Schmidt distance should be preferred over the approximate Log-Hilbert-Schmidt inner product and, computationally, it should be preferred over the approximate affine-invariant Riemannian distance. Numerical experiments on image classification demonstrate significant improvements of the infinite-dimensional formulation over the finite-dimensional counterpart. Given the numerous applications of covariance matrices in many areas of mathematics, statistics, and machine learning, just to name a few, we expect that the infinite-dimensional covariance operator formulation presented here will have many more applications beyond those in computer vision.  Read more... (not yet rated) 0 with reviews - Be the first.

## Find a copy in the library

Finding libraries that hold this item...

## Details

Genre/Form: Electronic books Print version: Document, Internet resource Internet Resource, Computer File Quang Minh Hà; Vittorio Murino Find more information about: Quang Minh Hà Vittorio Murino 9781681730141 1681730146 1012748003 1 online resource (xiii, 156 pages) : illustrations. Part I. Covariance matrices and applications -- 1. Data representation by covariance matrices -- 1.1 Covariance matrices for data representation -- 1.2 Statistical interpretation -- 2. Geometry of SPD matrices -- 2.1 Euclidean distance -- 2.2 Interpretations and motivations for the different invariances -- 2.3 Basic Riemannian geometry -- 2.4 Affine-invariant Riemannian metric on SPD matrices -- 2.4.1 Connection with the Fisher-Rao metric -- 2.5 Log-Euclidean metric -- 2.5.1 Log-Euclidean distance as an approximation of the affine-invariant Riemannian distance -- 2.5.2 Log-Euclidean distance as a Riemannian distance -- 2.5.3 Log-Euclidean vs. Euclidean -- 2.6 Bregman divergences -- 2.6.1 Log-determinant divergences -- 2.6.2 Connection with the Rényi and Kullback-Leibler divergences -- 2.7 Alpha-Beta Log-Det divergences -- 2.8 Power Euclidean metrics -- 2.9 Distances and divergences between empirical covariance matrices -- 2.10 Running time comparison -- 2.11 Summary -- 3. Kernel methods on covariance matrices -- 3.1 Positive definite kernels and reproducing kernel Hilbert spaces -- 3.2 Positive definite kernels on SPD matrices -- 3.2.1 Positive definite kernels with the Euclidean metric -- 3.2.2 Positive definite kernels with the log-Euclidean metric -- 3.2.3 Positive definite kernels with the symmetric Stein divergence -- 3.2.4 Positive definite kernels with the affine-invariant Riemannian metric -- 3.3 Kernel methods on covariance matrices -- 3.4 Experiments on image classification -- 3.4.1 Datasets -- 3.4.2 Results -- 3.5 Related approaches -- Part II. Covariance operators and applications -- 4. Data representation by covariance operators -- 4.1 Positive definite kernels and feature maps -- 4.2 Covariance operators in RKHS -- 4.3 Data representation by RKHS covariance operators -- 5. Geometry of covariance operators -- 5.1 Hilbert-Schmidt distance -- 5.2 Riemannian distances between covariance operators -- 5.2.1 The affine-invariant Riemannian metric -- 5.2.2 Log-Hilbert-Schmidt metric -- 5.3 Infinite-dimensional alpha log-determinant divergences -- 5.4 Summary -- 6. Kernel methods on covariance operators -- 6.1 Positive definite kernels on covariance operators -- 6.1.1 Kernels defined using the Hilbert-Schmidt metric -- 6.1.2 Kernels defined using the log-Hilbert-Schmidt metric -- 6.2 Two-layer kernel machines -- 6.3 Approximate methods -- 6.3.1 Approximate log-Hilbert-Schmidt distance and approximate affine-invariant Riemannian distance -- 6.3.2 Computational complexity -- 6.3.3 Approximate log-Hilbert-Schmidt inner product -- 6.3.4 Two-layer kernel machine with the approximate log-Hilbert-Schmidt distance -- 6.3.5 Case study: approximation by Fourier feature maps -- 6.4 Experiments in image classification -- 6.5 Summary -- 7. Conclusion and future outlook -- A. Supplementary technical information -- Mean squared errors for empirical covariance matrices -- Matrix exponential and principal logarithm Fréchet derivative -- The quasi-random Fourier features -- Low-discrepancy sequences -- The Gaussian case -- Proofs of several mathematical results -- Bibliography -- Authors' biographies. Synthesis lectures on computer vision, # 13. Hà Quang Minh, Vittorio Murino.

### Abstract:

Presents an overview of the {\it finite-dimensional covariance matrix} representation approach of images, along with its statistical interpretation. In particular, the book discusses the various  Read more...

## Reviews

User-contributed reviews

Be the first.

## Similar Items

### Related Subjects:(17)

Confirm this request

You may have already requested this item. Please select Ok if you would like to proceed with this request anyway.

### Primary Entity

<http://www.worldcat.org/oclc/1012748003> # Covariances in computer vision and machine learning
a schema:MediaObject, schema:Book, schema:CreativeWork ;
library:oclcnum "1012748003" ;
library:placeOfPublication <http://id.loc.gov/vocabulary/countries/cau> ;
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Thing/visual_object_recognition> ; # visual object recognition
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Topic/computers_general> ; # COMPUTERS / General
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Thing/bregman_divergences> ; # Bregman divergences
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Thing/positive_definite_operators> ; # positive definite operators
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Thing/logeuclidean_distance> ; # LogEuclidean distance
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Thing/affine_invariant_riemannian_distance> ; # affine-invariant Riemannian distance
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Thing/hilbert_schmidt_operators> ; # Hilbert-Schmidt operators
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Thing/log_hilbert_schmidt_distance> ; # Log-Hilbert-Schmidt distance
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Thing/kernel_methods_on_riemannian_manifolds> ; # kernel methods on Riemannian manifolds
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Thing/riemannian_manifolds> ; # Riemannian manifolds
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Topic/machine_learning_mathematical_models> ; # Machine learning--Mathematical models
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Topic/computer_vision_mathematical_models> ; # Computer vision--Mathematical models
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Thing/convex_cone> ; # convex cone
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Thing/image_classification> ; # image classification
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Thing/infinitedimensional_covariance_operators> ; # infinitedimensional covariance operators
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Thing/covariance_descriptors_in_computer_vision> ; # covariance descriptors in computer vision
schema:about <http://experiment.worldcat.org/entity/work/data/4617072560#Thing/positive_definite_matrices> ; # positive definite matrices
schema:author <http://experiment.worldcat.org/entity/work/data/4617072560#Person/ha_quang_minh_1977> ; # Quang Minh Hà
schema:author <http://experiment.worldcat.org/entity/work/data/4617072560#Person/murino_vittorio> ; # Vittorio Murino
schema:bookFormat schema:EBook ;
schema:datePublished "2018" ;
schema:description "Part I. Covariance matrices and applications -- 1. Data representation by covariance matrices -- 1.1 Covariance matrices for data representation -- 1.2 Statistical interpretation -- 2. Geometry of SPD matrices -- 2.1 Euclidean distance -- 2.2 Interpretations and motivations for the different invariances -- 2.3 Basic Riemannian geometry -- 2.4 Affine-invariant Riemannian metric on SPD matrices -- 2.4.1 Connection with the Fisher-Rao metric -- 2.5 Log-Euclidean metric -- 2.5.1 Log-Euclidean distance as an approximation of the affine-invariant Riemannian distance -- 2.5.2 Log-Euclidean distance as a Riemannian distance -- 2.5.3 Log-Euclidean vs. Euclidean -- 2.6 Bregman divergences -- 2.6.1 Log-determinant divergences -- 2.6.2 Connection with the Rényi and Kullback-Leibler divergences -- 2.7 Alpha-Beta Log-Det divergences -- 2.8 Power Euclidean metrics -- 2.9 Distances and divergences between empirical covariance matrices -- 2.10 Running time comparison -- 2.11 Summary -- 3. Kernel methods on covariance matrices -- 3.1 Positive definite kernels and reproducing kernel Hilbert spaces -- 3.2 Positive definite kernels on SPD matrices -- 3.2.1 Positive definite kernels with the Euclidean metric -- 3.2.2 Positive definite kernels with the log-Euclidean metric -- 3.2.3 Positive definite kernels with the symmetric Stein divergence -- 3.2.4 Positive definite kernels with the affine-invariant Riemannian metric -- 3.3 Kernel methods on covariance matrices -- 3.4 Experiments on image classification -- 3.4.1 Datasets -- 3.4.2 Results -- 3.5 Related approaches --"@en ;
schema:description "Covariance matrices play important roles in many areas of mathematics, statistics, and machine learning, as well as their applications. In computer vision and image processing, they give rise to a powerful data representation, namely the covariance descriptor, with numerous practical applications. In this book, we begin by presenting an overview of the finite-dimensional covariance matrix representation approach of images, along with its statistical interpretation. In particular, we discuss the various distances and divergences that arise from the intrinsic geometrical structures of the set of Symmetric Positive Definite (SPD) matrices, namely Riemannian manifold and convex cone structures. Computationally, we focus on kernel methods on covariance matrices, especially using the Log-Euclidean distance. We then show some of the latest developments in the generalization of the finite dimensional covariance matrix representation to the infinite-dimensional covariance operator representation via positive definite kernels. We present the generalization of the affine-invariant Riemannian metric and the Log-Hilbert-Schmidt metric, which generalizes the Log-Euclidean distance. Computationally, we focus on kernel methods on covariance operators, especially using the Log-Hilbert-Schmidt distance. Specifically, we present a two-layer kernel machine, using the Log-Hilbert-Schmidt distance and its finite-dimensional approximation, which reduces the computational complexity of the exact formulation while largely preserving its capability. Theoretical analysis shows that, mathematically, the approximate Log-Hilbert-Schmidt distance should be preferred over the approximate Log-Hilbert-Schmidt inner product and, computationally, it should be preferred over the approximate affine-invariant Riemannian distance. Numerical experiments on image classification demonstrate significant improvements of the infinite-dimensional formulation over the finite-dimensional counterpart. Given the numerous applications of covariance matrices in many areas of mathematics, statistics, and machine learning, just to name a few, we expect that the infinite-dimensional covariance operator formulation presented here will have many more applications beyond those in computer vision."@en ;
schema:exampleOfWork <http://worldcat.org/entity/work/id/4617072560> ;
schema:genre "Electronic books"@en ;
schema:inLanguage "en" ;
schema:isPartOf <http://worldcat.org/issn/2153-1064> ; # Synthesis lectures on computer vision ;
schema:isSimilarTo <http://worldcat.org/entity/work/data/4617072560#CreativeWork/> ;
schema:name "Covariances in computer vision and machine learning"@en ;
schema:productID "1012748003" ;
schema:url <http://dx.doi.org/10.2200/S00801ED1V01Y201709COV011> ;
schema:workExample <http://dx.doi.org/10.2200/S00801ED1V01Y201709COV011> ;
schema:workExample <http://worldcat.org/isbn/9781681730141> ;
wdrs:describedby <http://www.worldcat.org/title/-/oclc/1012748003> ;
.

### Related Entities

<http://experiment.worldcat.org/entity/work/data/4617072560#Person/ha_quang_minh_1977> # Quang Minh Hà
a schema:Person ;
schema:birthDate "1977" ;
schema:familyName "" ;
schema:givenName "Quang Minh" ;
schema:name "Quang Minh Hà" ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Person/murino_vittorio> # Vittorio Murino
a schema:Person ;
schema:familyName "Murino" ;
schema:givenName "Vittorio" ;
schema:name "Vittorio Murino" ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Thing/affine_invariant_riemannian_distance> # affine-invariant Riemannian distance
a schema:Thing ;
schema:name "affine-invariant Riemannian distance" ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Thing/bregman_divergences> # Bregman divergences
a schema:Thing ;
schema:name "Bregman divergences" ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Thing/covariance_descriptors_in_computer_vision> # covariance descriptors in computer vision
a schema:Thing ;
schema:name "covariance descriptors in computer vision" ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Thing/hilbert_schmidt_operators> # Hilbert-Schmidt operators
a schema:Thing ;
schema:name "Hilbert-Schmidt operators" ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Thing/image_classification> # image classification
a schema:Thing ;
schema:name "image classification" ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Thing/infinitedimensional_covariance_operators> # infinitedimensional covariance operators
a schema:Thing ;
schema:name "infinitedimensional covariance operators" ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Thing/kernel_methods_on_riemannian_manifolds> # kernel methods on Riemannian manifolds
a schema:Thing ;
schema:name "kernel methods on Riemannian manifolds" ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Thing/log_hilbert_schmidt_distance> # Log-Hilbert-Schmidt distance
a schema:Thing ;
schema:name "Log-Hilbert-Schmidt distance" ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Thing/logeuclidean_distance> # LogEuclidean distance
a schema:Thing ;
schema:name "LogEuclidean distance" ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Thing/positive_definite_matrices> # positive definite matrices
a schema:Thing ;
schema:name "positive definite matrices" ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Thing/positive_definite_operators> # positive definite operators
a schema:Thing ;
schema:name "positive definite operators" ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Thing/riemannian_manifolds> # Riemannian manifolds
a schema:Thing ;
schema:name "Riemannian manifolds" ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Thing/visual_object_recognition> # visual object recognition
a schema:Thing ;
schema:name "visual object recognition" ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Topic/computer_vision_mathematical_models> # Computer vision--Mathematical models
a schema:Intangible ;
schema:name "Computer vision--Mathematical models"@en ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Topic/computers_general> # COMPUTERS / General
a schema:Intangible ;
schema:name "COMPUTERS / General"@en ;
.

<http://experiment.worldcat.org/entity/work/data/4617072560#Topic/machine_learning_mathematical_models> # Machine learning--Mathematical models
a schema:Intangible ;
schema:name "Machine learning--Mathematical models"@en ;
.

<http://worldcat.org/entity/work/data/4617072560#CreativeWork/>
a schema:CreativeWork ;
schema:description "Print version:" ;
schema:isSimilarTo <http://www.worldcat.org/oclc/1012748003> ; # Covariances in computer vision and machine learning
.

<http://worldcat.org/isbn/9781681730141>
a schema:ProductModel ;
schema:isbn "1681730146" ;
schema:isbn "9781681730141" ;
.

<http://worldcat.org/issn/2153-1064> # Synthesis lectures on computer vision ;
a bgn:PublicationSeries ;
schema:hasPart <http://www.worldcat.org/oclc/1012748003> ; # Covariances in computer vision and machine learning
schema:issn "2153-1064" ;
schema:name "Synthesis lectures on computer vision ;" ;
schema:name "Synthesis lectures on computer vision," ;
.

Content-negotiable representations