ODYSSEY 2004 - The Speaker and Language Recognition Workshop May 31 - June 3, 2004 Toledo, Spain

On the Use of Quality Measures for Text-Independent Speaker Recognition

D. Garcia-Romero, J. Fierrez-Aguilar, Joaquin Gonzalez-Rodriguez, Javier Ortega-Garcia

Speech and Signal Processing Group (ATVS), Department of Audiovisual Engineering and Communications (DIAC), Technical University of Madrid (UPM), Spain

The use of quality information on automatic recognition systems is studied. From an apparent definition of what constitutes a quality measure, a framework for the successful exploitation of the quality information is derived. Potential applications are also introduced at different phases of the recognition process, namely: enrollment, scoring and multi-level fusion stages. Traditional likelihood scoring stage is further developed providing guidelines for the practical application of the proposed ideas. Preliminary experiments corroborate the benefits of the proposed quality-guided recognition approach. In particular, a frame-level quality measure meeting a goodness criterion based on deviation from the fundamental frequency is used, obtaining encouraging initial results.

Full Paper

Bibliographic reference.  Garcia-Romero, D. / Fierrez-Aguilar, J. / Gonzalez-Rodriguez, Joaquin / Ortega-Garcia, Javier (2004): "On the use of quality measures for text-independent speaker recognition", In ODYS-2004, 105-110.