Sixth International Conference on Spoken Language Processing
(ICSLP 2000)

Beijing, China
October 16-20, 2000

Phonetic Consistency in Spanish for Pin-Based Speaker Verification System

J. Ortega-Garcia (1), J. G. Rodriguez (1), D. T. Merino (2)

(1) EUIT Telecomunicación, Universidad Politécnica de Madrid, Spain
(2) Telefónica Investigación y Desarrollo, Madrid, Spain

The use of uttered Personal Identification Numbers (PIN) is a well-suited approach for person identification through voice in real applications. In this paper, speaker verification with short 4-digit strings, in a pragmatic perspective where very few utterances for training are available, is accomplished. The problem here arises due to the small quantity of voice available in short PIN utterances. Furthermore, it has to be taken into account the specificity of Spanish in this task, as digit strings are not uttered in a isolated digit-by-digit basis, but mentally grouped without constraints, and read as whole figures, with varying groups for different utterances of the same PIN. This specific factor induces high dependency on the phonetic contents of the PIN, and complicates considerably the design of text-dependent systems. Considering this, a textindependent GMM speaker verification system, including ‘nearest reference speaker’ and ‘universal background model’ score normalization, together with CMN channel compensation, has been evaluated over a specific PIN database, where different training conditions (phonetic dependent/independent) are tested.


Full Paper

Bibliographic reference.  Ortega-Garcia, J. / Rodriguez, J. G. / Merino, D. T. (2000): "Phonetic consistency in Spanish for pin-based speaker verification system", In ICSLP-2000, vol.2, 262-265.