Block-diagonal covariance joint subspace tying and model compensation for noise robust automatic speech recognition
US7729909B2 · kind B2 · utility
10Cited by
25References
29Claims
0Family size
Assignee
Inventors
Key dates
| Filing date | Mar 6, 2006 |
| Grant date | Jun 1, 2010 |
| Priority date | — |
| Expiry date | Apr 1, 2029 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG10L15/065
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
Model compression is combined with model compensation. Model compression is needed in embedded ASR to reduce the size and the computational complexity of compressed models. Model-compensation is used to adapt in real-time to changing noise environments. The present invention allows for the design of smaller ASR engines (memory consumption reduced to up to one-sixth) with reduced impact on recognition accuracy and/or robustness to noises.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.