System and method for distributed speech recognition
View Patent ↗A system and method for distributed speech recognition is provided. A prompt is provided to a caller during a call. One or more audio responses are received from the caller in response to the prompt. Distributed speech recognition is performed on the audio responses by providing a non-overlapping section of a main grammar to each of a plurality of secondary recognizers for each audio response. Speech recognition is performed on the audio responses by each of the secondary recognizers using the non-overlapping section of the main grammar associated with that secondary recognizer. A new grammar is generated based on results of the speech recognition from each of the secondary recognizers. Further speech recognition is performed on the audio responses against the new grammar and a further prompt is selected for providing to the caller based on results of the distributed speech recognition.
1. A system for distributed speech recognition, comprising:
a prompt provided to a caller during a call;
a receipt module to receive one or more audio responses from the caller in response to the prompt;
a distributed speech recognition module to perform distributed speech recognition on the audio responses, comprising:
a non-overlapping section of a main grammar provided to each of a plurality of secondary recognizers for each audio response;
a speech recognition module to perform speech recognition on the audio responses by each of the secondary recognizers using the non-overlapping section of the main grammar associated with that secondary recognizer;
a grammar module to provide a main grammar template to a main recognizer;
a grammar construction module to construct a new grammar based on the main grammar template using speech recognition search results from each of the secondary recognizers; and
a further speech recognition module to perform further speech recognition on the audio responses against the new grammar; and
a further prompt provided to the caller based on results of the distributed speech recognition.
2. A system according to claim 1 , further comprising:
a display module to display the results of the distributed speech recognition to an agent participating in the call.
3. A system according to claim 2 , further comprising:
a modification module to receive from the agent one or more modifications to the distributed speech recognition results.
4. A system according to claim 1 , further comprising:
an assignment module to assign to each secondary recognizer an independent grammar.
5. A system according to claim 4 , wherein the independent grammars comprise language-specific grammars.
6. A system according to claim 1 , further comprising:
a grammar formation module to form the new grammar, comprising:
a result module to determine the results of the speech recognition performed by each of the secondary recognizers; and
a result selection module to identify an n number of most likely search results from the results of each secondary recognizer.
7. A system according to claim 1 , further comprising:
a grammar module to utilize the new grammar during the remainder of the call.
8. A system according to claim 1 , further comprising:
a performance module to simultaneously perform the speech recognition by each of the secondary recognizers.
9. A system according to claim 1 , further comprising:
an interfacing module to interface the main recognizer to each of the secondary recognizers.
10. A method for distributed speech recognition, comprising:
transmitting a prompt to a caller during a call;
receiving one or more audio responses from the caller in response to the prompt;
performing distributed speech recognition on the audio responses, comprising:
providing a non-overlapping section of a main grammar to each of a plurality of secondary recognizers for each audio response;
performing speech recognition on the audio responses by each of the secondary recognizers using the non-overlapping section of the main grammar associated with that secondary recognizer;
providing a main grammar template to a main recognizer;
constructing a new grammar based on the main grammar template using speech recognition search results from each of the secondary recognizers; and
performing further speech recognition on the audio responses against the new grammar; and
selecting a further prompt for providing to the caller based on results of the distributed speech recognition.
11. A method according to claim 10 , further comprising:
displaying the results of the distributed speech recognition to an agent participating in the call.
12. A method according to claim 11 , further comprising:
receiving from the agent one or more modifications to the distributed speech recognition results.
13. A method according to claim 10 , further comprising:
assigning each secondary recognizer an independent grammar.
14. A method according to claim 13 , wherein the independent grammars comprise language-specific grammars.
15. A method according to claim 10 , further comprising:
forming the new grammar, comprising:
determining the results of the speech recognition performed by each of the secondary recognizers; and
identifying an n number of most likely search results from the results of each secondary recognizer.
16. A method according to claim 10 , further comprising:
utilizing the new grammar during the remainder of the call.
17. A method according to claim 10 , further comprising:
simultaneously performing the speech recognition by each of the secondary recognizers.
18. A method according to claim 10 , further comprising:
interfacing the main recognizer to each of the secondary recognizers.