Adaptive Biometric Systems Recent Advances and Challenges

This interdisciplinary volume presents a detailed overview of the latest advances and challenges remaining in the field of adaptive biometric systems. A broad range of techniques are provided from an international selection of pre-eminent authorities, collected together under a unified taxonomy and designed to be applicable to any pattern recognition system. Features: presents a thorough introduction to the concept of adaptive biometric systems; reviews systems for adaptive face recognition that perform self-updating of facial models using operational (unlabeled) data; describes a novel semi-supervised training strategy known as fusion-based co-training; examines the characterization and recognition of human gestures in videos; discusses a selection of learning techniques that can be applied to build an adaptive biometric system; investigates procedures for handling temporal variance in facial biometrics due to aging; proposes a score-level fusion scheme for an adaptive multimodal biometric system.

Introduction to Adaptive Biometric Systems.- Context-Sensitive Self-Updating for Adaptive Face Recognition.- Handling Session Mismatch by Semi-Supervised Based Co-Training Scheme.- A Hybrid CRF/HMM for One-Shot Gesture Learning.- An Online Learning-Based Adaptive Biometric System.- Adaptive Facial Recognition Under Aging Effect.- An Adaptive Score Level Fusion Scheme for Multimodal Biometric Systems.

Dr. Ajita Rattani is a post-doctoral fellow in the Integrated Pattern Recognition and Biometrics (i-PRoBe) lab at Michigan State University, East Lansing, MI, USA. Dr. Fabio Roli is a professor of computer engineering and the Director of the Pattern Recognition and Applications (PRA) lab at the University of Cagliari, Italy. Dr. Eric Granger is a professor in the Department of Automated Manufacturing Engineering and the Director of the Laboratory for Imagery, Vision and Artificial Intelligence at the École de technologie supérieure (ÉTS), Montréal, QC, Canada.