Face Recognition using Three-Dimensional and Multimodal Images

The great attention received by face recognition is motivated not only by the fundamental challenges posed by the problem, but also by numerous practical applications. Although current automatic recognition systems have reached a certain level of reliability, their success is limited by the conditions imposed by many practical applications. The recent improvements in three-dimensional acquisition devices allows the implementation of three-dimensional face recognition systems. Some devices incorporate a color digital camera that allows the acquisition of multimodal 2D+3D images, which can be exploited to design more reliable systems. This work investigates face recognition strategies based on the analysis of three-dimensional and multimodal images. The topics addressed include 3D face face detection and normalization, geometric features, information fusion strategies and recognition of partially occluded faces.

Autorentext

Claudio Cusano is a research fellow of the Imaging and Vision Laboratory of the University of Milano-Bicocca, where he took his PhD in Computer Science. Since 2001 he has been a fellow of the the ITC Institute of the Italian National Research Council. The main topics of his research concern computer vision, pattern recognition, and biometrics.