System and method for performing distributed speech recognition

• US 9,224,389 B2
• Filed: 08/26/2013
• Issued: 12/29/2015
• Est. Priority Date: 03/29/2007
• Status: Active Grant

• Current Assignee: Intellisist, Inc. dba Spoken Communications (Avaya Incorporated)
• Original Assignee: Intellisist, Inc. dba Spoken Communications (Avaya Incorporated)
• Inventors: Odinak, Gilad
• Primary Examiner(s): Neway, Samuel G
• Application Number:
  US14/010,501
• Publication Number:
  US 20130346080A1
• Time in Patent Office: 855 Days
• US Class Current: 1/1
• CPC Class Codes: G10L 13/08 Text analysis or generation of parameters for speech synthesis out of text, e.g. grapheme to phoneme translation, prosody generation or stress or intonation determination
• CPC Class Codes: G10L 15/18 using natural language modelling
• CPC Class Codes: G10L 15/193 Formal grammars, e.g. finite state automata, context free grammars or word networks
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
• CPC Class Codes: G10L 15/32 Multiple recognisers used in sequence or in parallel; Score combination systems therefor, e.g. voting systems
• CPC Class Codes: G10L 15/34 Adaptation of a single recogniser for parallel processing, e.g. by use of multiple processors or cloud computing
• CPC Class Codes: H04M 2201/38 Displays
• CPC Class Codes: H04M 2201/40 using speech recognition
• CPC Class Codes: H04M 3/51 Centralised call answering arrangements requiring operator intervention , e.g. call or contact centers for telemarketing
• CPC Class Codes: H04M 3/5166 in combination with interactive voice response systems or voice portals, e.g. as front-ends
• CPC Class Codes: H04M 3/5183 Call or contact centers with computer-telephony arrangements

Abstract

A system and method for performing distributed speech recognition is provided. Audio data is received on a main recognizer and on each of a plurality of secondary recognizers. Secondary grammars are transmitted to each of the secondary recognizers. The secondary recognizers each perform speech recognition on the audio data using the secondary grammar for that secondary recognizer. A new grammar is constructed based on results of the speech recognition by each of the secondary recognizers. The main recognizer performs speech recognition on the audio data using the new grammar.

Claims

1. A system for performing distributed speech recognition, comprising:

a telephone to provide audio data during an incoming call;

a message server comprising memory, an input port to receive the audio data, wherein the message server is configured to;

to transmit a new call message to a main recognizer;

to initiate a new speech recognition session for the main recognizer;

to transmit the new call message from the main recognizer to each of a plurality of secondary recognizers;

to initiate a new speech recognition session for each of the secondary recognizers;

to transmit a main grammar template to the main recognizer for each speech utterance received as the audio during the incoming call;

transmit to each of the secondary recognizers, secondary grammars;

to receive speech recognition results on the audio data by each of the secondary recognizers using the secondary grammar for that secondary recognizer;

to receive from the main recognizer a new grammar based on the main grammar template using results of the speech recognition by each of the secondary recognizers; and

to receive by the main recognizer, speech recognition results of the audio data using the new grammar.

2. A system according to claim 1, further comprising:

a grammar reference module to transmit a set of secondary grammar references to the main recognizer for each speech utterance received during the incoming call;

a reference transmission module to transmit from the main recognizer, one of the secondary grammar references to each of the secondary recognizers; and

a grammar identification module to identify for each secondary recognizer, the secondary grammar based on the secondary grammar reference transmitted to that secondary recognizer.

3. A system according to claim 2, further comprising:

a grammar application module to apply the secondary grammars to each speech utterance received during the incoming call; and

a recognizer to identify parts of speech within the speech utterance that match the secondary grammars as results of the speech recognition performed by the secondary recognizers.

4. A system according to claim 1, wherein the secondary grammars comprise at least one of independent secondary grammars and non-overlapping secondary grammars.

5. A system according to claim 4, wherein the independent secondary grammars comprise language-specific grammars.

6. A system according to claim 1, further comprising:

a result module to determine results from the speech recognition performed by the main recognizer; and

a display to provide the results to a user for at least one of modifying the speech recognition results and selecting a prompt for providing to the user.

7. A system according to claim 6, wherein the prompt is selected based on the results of the speech recognition performed by the main recognizer.

8. A system according to claim 1, further comprising:

a template module to transmit a secondary grammar template to each of the secondary recognizers;

a reference module to transmit grammar references to further recognizers;

a further speech recognition module to perform speech recognition by each of the further recognizers based on grammars associated with the grammar references; and

a further module to transmit results of the speech recognition by the further recognizers to the secondary recognizers and constructing new secondary grammars using the results of the further recognizers.

9. A method for performing distributed speech recognition, comprising:

receiving audio data via a telephone during an incoming call;

transmitting via a server a new call message to a main recognizer, wherein the server comprises memory, an input port to receive the audio data, and an output port;

initiating a new speech recognition session for the main recognizer;

transmitting the new call message from the main recognizer to each of a plurality of secondary recognizers and initiating a new speech recognition session for each of the secondary recognizers;

transmitting a main grammar template to the main recognizer for each speech utterance received as the audio during the incoming call;

transmitting to each of the secondary recognizers, secondary grammars;

receiving speech recognition results on the audio data by each of the secondary recognizers using the secondary grammar for that secondary recognizer;

receiving from the main recognizer a new grammar based on the main grammar template using the results of the speech recognition by each of the secondary recognizers; and

receiving by the main recognizer, speech recognition results of the audio data using the new grammar.

10. A method according to claim 9, further comprising:

transmitting a set of secondary grammar references to the main recognizer for each speech utterance received during the incoming call;

transmitting from the main recognizer, one of the secondary grammar references to each of the secondary recognizers; and

identifying for each secondary recognizer, the secondary grammar based on the secondary grammar reference transmitted to that secondary recognizer.

11. A method according to claim 10, further comprising:

applying the secondary grammars to each speech utterance received during the incoming call; and

identifying parts of speech within the speech utterance that match the secondary grammars as results of the speech recognition performed by the secondary recognizers.

12. A method according to claim 9, wherein the secondary grammars comprise at least one of independent secondary grammars and non-overlapping secondary grammars.

13. A method according to claim 12, wherein the independent secondary grammars comprise language-specific grammars.

14. A method according to claim 9, further comprising:

determining results from the speech recognition performed by the main recognizer; and

providing the results to a user for at least one of modifying the speech recognition results and selecting a prompt for providing to the user.

15. A method according to claim 14, wherein the prompt is selected based on the results of the speech recognition performed by the main recognizer.

16. A method according to claim 9, further comprising:

transmitting a secondary grammar template to each of the secondary recognizers;

transmitting grammar references to further recognizers;

performing speech recognition by each of the further recognizers based on grammars associated with the grammar references; and

transmitting results of the speech recognition by the further recognizers to the secondary recognizers and constructing new secondary grammars using the results of the further recognizers.