Lip Motion Analysis for a Person Authentication System under Near Infrared Illumination

Lip motion analysis can enhance a security system based on face recognition significantly. Examining speaker dependent lip movements and visually determining the speech content can prevent attacks using photographs or prerecorded video sequences. We present an approach for lip motion analysis under near infrared illumination that also covers the task of the localization of the mouth area. We investigate different appearance-based features to model lip motions with hidden Markov models (HMMs) and evaluate the classification accuracy with our own TUNIR data base, which we made available to the public for the future comparison of competing approaches.