Full text
Preview |
PDF
- Requires a PDF viewer, such as GSview, Xpdf or Adobe Acrobat Reader
Download (273kB) | Preview |
Glottal Source Cepstrum Coefficients Applied to NIST SRE 2010
Citation
Mazaira Fernández, Luis Miguel, Álvarez Marquina, Agustin 
ORCID: https://orcid.org/0000-0002-6106-6711, Gómez Vilda, Pedro ORCID: https://orcid.org/0000-0003-3283-378X, Martínez Olalla, Rafael ORCID: https://orcid.org/0000-0003-2336-9145 and Muñoz, Cristina
(2010).
Glottal Source Cepstrum Coefficients Applied to NIST SRE 2010.
In: "V Jornadas de Reconocimiento Biométrico de Personas JRBP10", 02/09/2010 - 03/09/2010, Huesca, España.
Description
| Title: | Glottal Source Cepstrum Coefficients Applied to NIST SRE 2010 |
|---|---|
| Author/s: |
|
| Item Type: | Presentation at Congress or Conference (Article) |
| Event Title: | V Jornadas de Reconocimiento Biométrico de Personas JRBP10 |
| Event Dates: | 02/09/2010 - 03/09/2010 |
| Event Location: | Huesca, España |
| Title of Book: | Actas de las V Jornadas de Reconocimiento Biométrico de Personas JRBP10 |
| Date: | 2010 |
| Subjects: | |
| Freetext Keywords: | Glotal Source, Speaker Characterization, Speaker Recognition, GMM, Speech production, NIST SRE 2010 |
| Faculty: | Facultad de Informática (UPM) |
| Department: | Arquitectura y Tecnología de Sistemas Informáticos |
| Creative Commons Licenses: | Recognition - No derivative works - Non commercial |
Abstract
Through the present paper, a novel feature set for speaker recognition based on glottal estimate information is presented. An iterative algorithm is used to derive the vocal tract and glottal source estimations from speech signal. In order to test the importance of glottal source information in speaker characterization, the novel feature set has been tested in the 2010 NIST Speaker Recognition Evaluation (NIST SRE10). The proposed system uses glottal estimate parameter templates and classical cepstral information to build a model for each speaker involved in the recognition process. ALIZE [1] open-source software has been used to create the GMM models for both background and target speakers. Compared to using mel-frequency cepstrum coefficients (MFCC), the misclassification rate for the NIST SRE 2010 reduced from 29.43% to 27.15% when glottal source features are used
More information
| Item ID: | 7905 |
|---|---|
| DC Identifier: | https://oa.upm.es/7905/ |
| OAI Identifier: | oai:oa.upm.es:7905 |
| Official URL: | http://jrbp10.unizar.es/ |
| Deposited by: | Memoria Investigacion |
| Deposited on: | 27 Jul 2011 11:35 |
| Last Modified: | 20 Apr 2016 16:51 |
