Customer Reviews

Biometric Person Recognition: Face, Speech and Fusion

5 star:		(0)
4 star:		(1)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

This book is aimed at engineers, scientists and graduate students. It provides an overview of research and results in the areas of information fusion, speaker and face recognition, with a specific focus on how it applies to biometrics. While it doesn't cover every possible area and not all the developments, it is a good introduction to the main concepts and approaches. In some ways it can be thought of as a time saver (instead of reading sets of disparate articles), which additionally provides interesting research results.

Amongst other things, the author shows that automatic face recognition can be performed quite successfully even if the topology of the face is lost. This is at first counter intuitive, but provides advantages in terms of robustness to face alignment problems, in addition to reducing pose and illumination issues. The author then expands the idea by showing that pose changes can be handled by transformations of statistical model parameters rather than image transformations. Even though more details and more elaborate explanations would have been helpful to people less versed in the area, this book is a useful addition to a library on biometrics.