Foreword

This book addresses a fundamental step in face recognition research answering, among other issues, the following questions: how to properly measure the distance between surfaces representing faces, what are the pros and contras of each algorithms and how they compare with each other, what are their computational costs. In this respect, this book represents a reference point for PhD students and researchers who want to start working not only at face recognition problems but also at other applications dealing with the recognition of three-dimensional shapes. The need for such a book was particularly evident when we presented to our multidisciplinary team of the High Polytechnic School the topic to be studied that was aimed at the development of a diagnostic tool of prenatal syndromes from three-dimensional ultrasound scans (SYN DIAG). A book, easy to use, putting order and organizing the scientific significance of similarity measures applied to face recognition problems was missing. This aspect was crucial to support the choice of measures to be selected and tested. Coming to the topic of the book, face recognition has several applications, including security, such as authentication and identification of suspects, and medical ones, such as corrective surgery and diagnosis. So, I think that this book is going to be a valuable tool for all scientists 'facing face'.

Similarity Measures for Face Recognition Face recognition has several applications, including security, such as (authentication and identification of device users and criminal suspects), and in medicine (corrective surgery and diagnosis). Facial recognition programs rely on algorithms that can compare and compute the similarity between two sets of images. This eBook explains some of the similarity measures used in facial recognition systems in a single volume. Readers will learn about various measures including Minkowski distances, Mahalanobis distances, Hansdorff distances, cosine-based distances, among other methods. The book also summarizes errors that may occur in face recognition methods. Computer scientists "facing face" and looking to select and test different methods of computing similarities will benefit from this book. The book is also useful tool for students undertaking computer vision courses.


Author(s):  
Gisela Widmer

The stationary monochromatic radiative transfer equation (RTE) is posed in five dimensions, with the intensity depending on both a position in a three-dimensional domain as well as a direction. For non-scattering radiative transfer, sparse finite elements [1, 2] have been shown to be an efficient discretization strategy if the intensity function is sufficiently smooth. Compared to the discrete ordinates method, they make it possible to significantly reduce the number of degrees of freedom N in the discretization with almost no loss of accuracy. However, using a direct solver to solve the resulting linear system requires O(N3) operations. In this paper, an efficient solver based on the conjugate gradient method (CG) with a subspace correction preconditioner is presented. Numerical experiments show that the linear system can be solved at computational costs that are nearly proportional to the number of degrees of freedom N in the discretization.


2015 ◽  
Vol 15 (2) ◽  
pp. 377-378 ◽  
Author(s):  
Estefania López Rodriguez ◽  
Rosario Garcia Jimenez ◽  
Marta Sanchez Aguilar ◽  
Julio Valencia Anguita ◽  
Javier Luis Simon

Sign in / Sign up

Export Citation Format

Share Document