Diagnostic techniques based on speech models
US11024327B2 · kind B2 · utility
Assignee
Inventor
Key dates
| Filing date | Mar 12, 2019 |
| Grant date | Jun 1, 2021 |
| Priority date | — |
| Expiry date | Oct 23, 2039 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG10L25/66
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
A speech model constructed from reference speech samples, which were produced by a subject at a first time while a physiological state of the subject was known, is obtained. The model includes (i) acoustic states exhibited in the reference speech samples and associated with respective local distance functions, and (ii) allowed transitions between the acoustic states. A test speech sample that was produced by the subject at a second time, while the physiological state of the subject was unknown, is received. Test-sample feature vectors that quantify acoustic features of different respective portions of the test speech sample are computed. Based on the local distance functions and the allowed transitions, the test speech sample is mapped to a minimum-distance sequence of the acoustic states. In response thereto, an output indicating the physiological state of the subject at the second time is generated. Other embodiments are also described.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.