Method and system of enabling intelligent and lightweight speech to text transcription through distributed environment
US8438025B2 · kind B2 · utility
Assignee
Inventors
Key dates
| Filing date | Jul 30, 2012 |
| Grant date | May 7, 2013 |
| Priority date | — |
| Expiry date | Jul 30, 2032 |
Classification
- Technology area (CPC H)Electricity
- CPC primaryH04M2201/60
- WIPO fieldTelecommunications
- WIPO sectorElectrical engineering
Abstract
A system includes at least one wireless client device, a service manager, and a plurality of voice transcription servers. The service manager includes a resource management service and a profile management service. The client device communicates the presence of a voice transcription task to the resource management service. The resource management service surveys the plurality of voice transcription servers and selects one voice transcription server based on a set of predefined criteria. The resource management service then communicated an address of the selected server to the profile management service, which then transmits a trained voice profile or default profile to the selected server. The address of the selected server is then sent to the client device, which then transmits an audio stream to the server. Finally, the selected server transcribes the audio stream to a textual format.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.