Distributed voice recognition system

• US 5,956,683 A
• Filed: 04/04/1996
• Issued: 09/21/1999
• Est. Priority Date: 12/22/1993
• Status: Expired due to Term

• Current Assignee: Qualcomm, Inc.
• Original Assignee: Qualcomm, Inc.
• Inventors: Chang, Chienchung, Jacobs, Paul E.
• Primary Examiner(s): Hudspeth, David R.
• Assistant Examiner(s): Opsasnick, Michael N.
• Application Number:
  US08/627,333
• Time in Patent Office: 1,265 Days
• Field of Search: 395/2.09, 395/704, 395/275, 395/270, 395/201, 395/251, 395/455, 395/411, 395/412, 395/413
• US Class Current: 704/270.1
• CPC Class Codes: G10L 15/02 Feature extraction for speech recognition; Selection of recognition unit
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
• CPC Class Codes: G10L 19/00 Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis in musical instruments G10H

Abstract

A voice recognition system having a feature extraction apparatus is located in a remote station. The feature extraction apparatus extracts features from an input speech frame and then provides the extracted features to a central processing station. In the central processing station, the features are provided to a word decoder which determines the syntax of the input speech frame.

Claims

1. A remote station for use in a mobile communications system, comprising:

feature extraction means located at the remote station for receiving a frame of speech samples and extracting a set of parameters for speech recognition;

first word decoder means for receiving said set of parameters and for extracting the meaning of said speech from said parameters in accordance with a small vocabulary; and

transmission means for wirelessly transmitting a set of parameters that cannot be decoded by said first word decoder means to a receiving station having second word decoder means for extracting the meaning of said speech from the transmitted parameters in accordance with a larger vocabulary.

2. The remote station of claim 1, further comprising a microphone for receiving an acoustical signal and for providing said acoustical signal to said feature extraction means.

3. The remote station of claim 1, further comprising transform means interposed between said feature extraction means and said transmitter means for receiving said set of parameters and for transforming said set of parameters into an alternative representation of said set of parameters in accordance with a predetermined transformation format.

4. The remote station of claim 1, wherein said set of parameters comprises linear prediction coefficients.

5. The remote station of claim 1, wherein said set of parameters comprises line spectral pair values.

6. The remote station of claim 3, wherein said set of parameters comprises linear prediction coefficients and said predetermined transformation format is a linear prediction coefficient to line spectral pair transformation.

7. The remote station of claim 1, further comprising receiver means for receiving a response signal in accordance with a speech recognition operation upon said frame of speech.

8. The remote station of claim 7, further comprising control means for receiving said response signal and for providing a control signal in accordance with said response signal.

9. The remote station of claim 1, further comprising a transform element interposed between and having an input coupled to an output of said feature extraction means and having an output coupled to an input of said transmission means.

10. The remote station of claim 7, further comprising a control element having an input coupled to said receiver output and having an output for providing a control signal in accordance with said response signal.

11. A central communications station for use in a mobile communications system, comprising:

a word decoder located at said central communications station for receiving a set of speech parameters from a remote station physically separated from said central communications station and communicating therewith by wireless communications means, for performing a speech recognition operation on said set of speech parameters using a regular vocabulary associated with said word decoder located at said central communications station, wherein said speech parameters are not recognizable by a local vocabulary associated with a word decoder located in a remote station; and

a signal generator for generating a response signal based on a result of said speech recognition operation.

12. The central communications station of claim 11, further comprising a receiver having an input for receiving a remote station speech parameter signal and for providing said remote station speech parameters to said word decoder means.

13. The central communications station of claim 11, further comprising control means having an input coupled to said word decoder output and having an output for providing a control signal.

14. A distributed voice recognition system, comprising:

a local word decoder located at a subscriber station for receiving extracted acoustic features of a first frame of speech samples and for decoding said acoustic features in accordance with a small vocabulary; and

a remote word decoder located at a central processing station physically separated from said subscriber station for receiving extracted acoustic features of a second frame and for decoding, in accordance with a regular vocabulary, larger than said small vocabulary, said acoustic features of said second frame which cannot be decoded by said local word decoder.

15. The system of claim 14 further comprising a preprocessor for extracting said acoustic features of said frame of speech samples in accordance with a predetermined feature extraction format and for providing said acoustic features.

16. The system of claim 15 wherein said acoustic features are linear predictive coding (LPC) based parameters.

17. The system of claim 15 wherein said acoustic features are cepstral coefficients.

18. The system of claim 15 wherein said preprocessor comprises a voice coder (vocoder).

19. The system of claim 18 wherein said vocoder is a code excited linear prediction (CELP) vocoder.

20. The system of claim 18 wherein said vocoder is a linear predictive coding (LPC) based vocoder.

21. The system of claim 18 wherein said vocoder is a multi-band excitation (MBE) based vocoder.

22. The system of claim 18 wherein said vocoder is an ADPCM vocoder.

23. The system of claim 14 further comprising:

a transform element located at said subscriber station for receiving said acoustic features and for converting said acoustic features into transformed features in accordance with a predetermined transform format, said transformed features transmitted through a communication channel to said central processing station; and

an inverse transform element located at said central processing station for receiving said transformed features and for converting said transformed features into estimated acoustic features in accordance with a predetermined inverse transform format, and for providing said estimated acoustic features to said remote word decoder.

24. The system of claim 23 wherein said acoustic features are linear predictive coding (LPC) based parameters, said predetermined transform format converts said LPC based parameters to line spectral pair (LSP) frequencies, and said inverse transform format converts said LSP frequencies into LPC based parameters.

25. The system of claim 14 wherein said local word decoder performs acoustic pattern matching based on a hidden Markov model (HMM).

26. The system of claim 14 wherein said remote word decoder performs acoustic pattern matching based on a hidden Markov model (HMM).

27. The system of claim 14 wherein said local word decoder performs acoustic pattern matching based on Dynamic Time Warping (DTW).

28. The system of claim 14 wherein said remote word decoder performs acoustic pattern matching based on Dynamic Time Warping (DTW).

29. The system of claim 14 wherein said subscriber station communicates with said central processing station by wireless communication means.

30. The system of claim 29 wherein said wireless communication means comprises a CDMA communication system.