SPEECH UNDERSTANDING METHOD AND SYSTEM

• US 20090234651A1
• Filed: 03/12/2009
• Published: 09/17/2009
• Est. Priority Date: 03/12/2008
• Status: Active Grant

• Current Assignee: Appy Risk Technologies Limited (Trak Global Group Ltd.)
• Original Assignee: Intelligent Mechatronic Systems Inc. (Trak Global Group Ltd.)
• Inventors: Miners, William Ben, Basir, Otman A.
• Granted Patent: US 8,364,486 B2
• Time in Patent Office: Days
• US Class Current: 704/254
• CPC Class Codes: G10L 15/02 Feature extraction for speech recognition; Selection of recognition unit
• CPC Class Codes: G10L 15/28 Constructional details of speech recognition systems
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
• CPC Class Codes: G10L 2015/025 Phonemes, fenemes or fenones being the recognition units

Abstract

A speech recognition system includes a mobile device and a remote server. The mobile device receives the speech from the user and extracts the features and phonemes from the speech. Selected phonemes and measures of uncertainty are transmitted to the server, which processes the phonemes for speech understanding and transmits a text of the speech (or the context or understanding of the speech) back to the mobile device.

Claims

1. A mobile device including:

a transducer for receiving audible speech;

a phonetic decomposition module extracting phonemes based upon the audible speech received by the transducer; and

a transmitter for sending the phonemes.

2. The mobile device of claim 1 further including a receiver for receiving information generated based upon the phonemes transmitted by the transmitter.

3. The mobile device of claim 1 further including a controller for determining whether the audible speech has a high recognition uncertainty.

4. The mobile device of claim 3 wherein the controller decides whether the transmitter will send supplemental, non-verbal information based upon the determination of whether the audible speech has high recognition uncertainty.

5. The mobile device of claim 4 wherein the supplemental, non-verbal information includes images of a mouth of a person who is speaking the audible speech.

6. The mobile device of claim 3 wherein the controller decides whether the transmitter will send the phonemes based upon the determination of whether the audible speech has high recognition uncertainty.

7. The mobile device of claim 3 wherein the mobile device processes the phonemes based upon a determination that the audible speech does not have high recognition uncertainty, and sends the phonemes based upon the determination that the audible speech has high recognition uncertainty.

8. A speech recognition system including the mobile device of claim 1 further including a server receiving the phonemes from the transmitter, the server processing the phonemes to determine context.

9. The speech recognition system of claim 8 wherein the server sends information to the mobile device based upon the context of the phonemes.

10. The speech recognition system of claim 9 wherein the server determines context based upon the phonemes and based upon supplemental, non-verbal information.

11. The speech recognition system of claim 10 wherein the supplemental, non-verbal information includes a current task.

12. The speech recognition system of claim 10 wherein the supplemental, non-verbal information includes historical behavior patterns.

13. The speech recognition system of claim 10 wherein the mobile device decides whether the transmitter will send supplemental, non-verbal information based upon a determination of whether the audible speech has high recognition uncertainty.

14. The speech recognition system of claim 13 wherein the supplemental, non-verbal information includes parameters of a mouth of a person who is speaking the audible speech.

15. A method for processing audible speech including the steps of:

acquiring a speech signal;

processing the speech signal to generate phonemes; and

transmitting the phonemes to a remote location.

16. A method for processing audible speech including the steps of:

receiving phonemes extracted from audible speech at a location remote from a person speaking the audible speech;

processing the phonemes to determine context; and

transmitting the information to the remote location based upon the context.