USING PITCH DURING SPEECH RECOGNITION POST-PROCESSING TO IMPROVE RECOGNITION ACCURACY

• US 20110144987A1
• Filed: 12/10/2009
• Published: 06/16/2011
• Est. Priority Date: 12/10/2009
• Status: Active Grant

• Current Assignee: General Motors LLC (General Motors Company)
• Original Assignee: General Motors LLC (General Motors Company)
• Inventors: Zhao, Xufang, Arun, Uma
• Granted Patent: US 9,484,027 B2
• Time in Patent Office: Days
• US Class Current: 704/233
• CPC Class Codes: G10L 15/20 Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech G10L21/02 takes precedence
• CPC Class Codes: G10L 2015/027 Syllables being the recognition units
• CPC Class Codes: G10L 25/15 the extracted parameters being formant information
• CPC Class Codes: G10L 25/90 Pitch determination of speech signals

Abstract

A method of automated speech recognition in a vehicle. The method includes receiving audio in the vehicle, pre-processing the received audio to generate acoustic feature vectors, decoding the generated acoustic feature vectors to produce at least one speech hypothesis, and post-processing the at least one speech hypothesis using pitch to improve speech recognition accuracy. The speech hypothesis can be accepted as recognized speech during post-processing if pitch is present in the received audio. Alternatively, a pitch count for the received audio can be determined, N-best speech hypotheses can be post-processed by comparing the pitch count to syllable counts associated with the speech hypotheses, and the speech hypothesis having a syllable count equal to the pitch count can be accepted as recognized speech.

Claims

1. A method of automated speech recognition in a vehicle, the method comprising the steps of:

(a) receiving audio in the vehicle;

(b) pre-processing the received audio to generate acoustic feature vectors;

(c) decoding the generated acoustic feature vectors to produce at least one speech hypothesis; and

(d) post-processing the at least one speech hypothesis using pitch to improve speech recognition accuracy.

2. The method of claim 1, wherein step (d) includes rejecting the speech hypothesis if pitch is not detected in the received audio.

3. The method of claim 1, wherein step (d) includes comparing syllable counts associated with speech hypotheses to a pitch count of the received audio, and accepting as recognized speech a particular speech hypothesis having a syllable count equal to the pitch count.

4. The method of claim 3, wherein the accepted speech hypothesis has the highest confidence score of multiple speech hypotheses having a syllable count equal to the pitch count.

5. A method of automated speech recognition in a vehicle, the method comprising the steps of:

(a) receiving audio in the vehicle;

(b) generating a plurality of speech hypotheses from the received audio;

(c) determining a pitch count for the received audio;

(d) identifying one of the speech hypotheses as having a symbol count equal to the pitch count; and

(e) accepting the identified speech hypothesis as recognized speech for the received audio.

6. A method of automated speech recognition in a vehicle, the method comprising the steps of:

(a) receiving audio in the vehicle;

(b) pre-processing the received audio to generate acoustic feature vectors;

(c) processing the generated acoustic feature vectors to produce N-best speech hypotheses and associated confidence scores and syllable counts;

(d) determining a pitch count for the received audio;

(e) post-processing the N-best speech hypotheses by comparing syllable counts associated with the speech hypotheses to the pitch count; and

(f) accepting as recognized speech a speech hypothesis of the N-best speech hypotheses having a syllable count equal to the pitch count.

7. The method of claim 6, wherein the accepted speech hypothesis has the highest confidence score of multiple speech hypotheses having a syllable count equal to the pitch count.