The University of Kent has been among the pioneers of biometrics research in the UK. They have generated novel ideas on intelligent agents in biometric processing, multimodal strategies for biometric systems and the use of biometric control to establish trust and confidence the processing of distributed electronic documents. Their experience in biometrics also extends to the collection of biometric data and database development, including multi-modal biometric databases, and the evaluation of practical systems as well as active involvement in international standardisation efforts in the field of biometrics. The University of Kent has been involved in the recognition of facial images for a number of years. In recent years the focus of our work has been to use facial recognition as a component in multimodal biometric systems; exploring strategies for its effective integration. Earlier work had explored the potential for combining voice and lip-movement information for speaker recognition. Recent work has been focused on the use of color features and color space combinations for face detection and segmentation as a first stage to biometric processing. In addition they are exploring the possibilities for face detection and recognition in fractal compressed domains. Here they attempt to exploit scale and transformation invariant properties of novel features extracted from the fractal domain to achieve more robust face recognition. The achievements of our research group has been rewarded by the allocation of a major grant (SRIF) to enhance our research infrastructure in the area of biometrics and this has enabled the acquisition of an advanced 3D camera system and a high performance video processing computing platform which is used for synthesis and recognition of 3D facial images.
UKE will lead WP2.4 Multi-modal Features and Fusion Research and contribute strongly to WP3.1 Performance and Security Testing.