ESCA Workshop on Automatic Speaker Recognition, Identification, and Verification

Martigny, Switzerland
April 7-9, 1994

The Rôle of the Reference Template in Speaker Verification

M. I. Hannah (1), A. T. Sapeluk (1), Robert I. Damper (2)

(1) Department of Electronic and Electrical Engineering, Dundee Institute of Technology, Dundee, Scotland, UK
(2) Department of Electronics and Computer Science, University of Southampton, England, UK

This paper compares 4 different ways of using initial ('training') recordings as reference data ('templates') to be compared with a test token for speaker verification. We show that merging the original training data into a single template - as in 2 of the 4 methods - incurs a loss in speaker discrimination relative to the remaining 2 methods which retain all the data. However, these techniques do have the advantage of reduced computational complexity in terms of both storage requirements and processing time for verification.

The best-performing of our methods used a decision rule in which 60% of the individual matches to each retained template had to satisfy a threshold criterion. However, statistical significance of the superiority of this implementation relative to the next-best, which took a mean of the 5 individual scores as a basis for its accept/reject decision, has not yet been established. This awaits further work with a larger database.

Full Paper

Bibliographic reference.  Hannah, M. I. / Sapeluk, A. T. / Damper, Robert I. (1994): "The rle of the reference template in speaker verification", In ASRIV-1994, 181-184.