Speaker verification and speaker identification based on eigenvoices
US6141644A · kind A · utility
Assignee
Inventors
Key dates
| Filing date | Sep 4, 1998 |
| Grant date | Oct 31, 2000 |
| Priority date | — |
| Expiry date | Sep 4, 2018 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG10L17/02
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
Speech models are constructed and trained upon the speech of known client speakers (and also impostor speakers, in the case of speaker verification). Parameters from these models are concatenated to define supervectors and a linear transformation upon these supervectors results in a dimensionality reduction yielding a low-dimensional space called eigenspace. The training speakers are then represented as points or distributions in eigenspace. Thereafter, new speech data from the test speaker is placed into eigenspace through a similar linear transformation and the proximity in eigenspace of the test speaker to the training speakers serves to authenticate or identify the test speaker.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.