Name
Abstract | ||
|---|---|---|
The biometric signature derived from the estimation of the power spectral density singularities of a speaker's glottal source is described in the present work. This consists in the collection of peak-trough profiles found in the spectral den- sity, as related to the biomechanics of the vocal folds. Sam- ples of parameter estimations from a set of 100 normo- phonic (pathology-free) speakers are produced. Mapping the set of speaker's samples to a manifold defined by Principal Component Analysis and clustering them by k-means in terms of the most relevant principal components shows the separation of speakers by gender. This means that the pro- posed signature conveys relevant speaker's meta- information, which may be useful in security and forensic applications for which contextual side information is con- sidered relevant. |
Year | Venue | Field |
|---|---|---|
2008 | European Signal Processing Conference | Vocal folds,Pattern recognition,Computer science,Side information,Speech recognition,Spectral density,Artificial intelligence,Gravitational singularity,Biometrics,Cluster analysis,Manifold,Principal component analysis |
DocType | ISSN | Citations |
Conference | 2219-5491 | 0 |
PageRank | References | Authors |
0.34 | 8 | 7 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| P. Gómez | 1 | 5 | 2.10 |
| R. Fernandez | 2 | 0 | 0.34 |
| A. Alvarez | 3 | 0 | 0.34 |
| Luis Miguel Mazaira-Fernández | 4 | 2 | 0.75 |
| V. Rodellar | 5 | 14 | 7.98 |
| R. Martínez | 6 | 29 | 13.46 |
| Cristina Muñoz-Mulas | 7 | 0 | 0.34 |