Abstract
© 2015 by John Wiley & Sons, Ltd. All rights reserved.Traditionally, user identification or authentication techniques have been based on what one possesses (such as access cards) or what one knows (such as passwords). In contrast, biometrics is based on what one is, such as the physiological or behavioral traits of the user. As human-computer interactions become more natural, there is an increasing demand for authentications based on biometrics. In this chapter, we explore multimodal biometrics as a viable means to take advantage of the diverse ways that a human interacts with the system. We survey a number of scenarios where multimodal biometrics are required. We then outline a new, hand-held system for identity verification that uses both face and voice characteristics of the user. In addition, experiments show that, with audio feedback and proper instruction, blind subjects can potentially use the technology for authentication and for performing secure transactions.