User Verification System

This project involves the application of pattern recognition techniques to establish the discriminative power of forensic biometric data. In particular, it concerns the verification of a person based on properties (features) of his/her voice, face, and/or iris. The verification decision is binary: either yes you are the person you claim to be or no you are not the person you claim to be.

A large portion of the effort will involve the establishment of databases with Web interfaces for voice prints, face images, and iris scans. This project will build on work initiated last year in a project from CS631 Pervasive Computing, see the paper below.

W. Baker, A. Evans, L. Jordan, and S. Pethe, “User verification system,” Proc. 8th Annual Mid-Atlantic Student Workshop on Programming Languages and Systems (MASPLAS'02), April 2002. A copy of this paper can be found on the Schedule page at MASPLAS'02.