System for improving speech intelligibility through high frequency compression

• US 8,086,451 B2
• Filed: 12/09/2005
• Issued: 12/27/2011
• Est. Priority Date: 04/20/2005
• Status: Active Grant

• Current Assignee: Malikie Innovations Limited (Key Patent Innovations Limited)
• Original Assignee: QNX Software Systems Company (Blackberry Limited)
• Inventors: Hetherington, Phillip A., Li, Xueman
• Primary Examiner(s): Wozniak, James S
• Assistant Examiner(s): He, Jialong
• Application Number:
  US11/298,053
• Publication Number:
  US 20060241938A1
• Time in Patent Office: 2,209 Days
• Field of Search: 704/201, 704/205, 704/225, 704500-501, 381/316
• US Class Current: 704/225
• CPC Class Codes: G10L 21/02 Speech enhancement, e.g. noise reduction or echo cancellation reducing echo effects in line transmission systems H04B3/20; echo suppression in hands-free telephones H04M9/08
• CPC Class Codes: G10L 21/0232 Processing in the frequency domain
• CPC Class Codes: G10L 21/0264 characterised by the type of parameter measurement, e.g. correlation techniques, zero crossing techniques or predictive techniques
• CPC Class Codes: G10L 21/0364 for improving intelligibility

Abstract

A speech enhancement system that improves the intelligibility and the perceived quality of processed speech includes a frequency transformer and a spectral compressor. The frequency transformer converts speech signals from the time domain to the frequency domain. The spectral compressor compresses a pre-selected portion of the high frequency band and maps the compressed high frequency band to a lower band limited frequency range.

Claims

1. A speech system that improves the intelligibility and quality of a processed speech, comprising:

a frequency transformer device that converts a speech signal into a spectrum of frequencies;

a spectral compressor device electrically coupled to the frequency transformer that compresses a pre-selected high frequency band of the speech signal and maps the compressed high frequency band to a lower band limited frequency range; and

a gain controller device that applies a variable gain to the compressed high frequency band in relation to a background noise level present in the speech signal, where the gain controller selects a level for the variable gain based on a slope of a noise floor present in the compressed high frequency band of the speech signal and a slope of a noise floor present in an uncompressed frequency portion of the speech signal.

2. The system of claim 1, where the frequency converter automatically converts the speech signal into its frequency spectrum in nearly real time.

3. The system of claim 1, where the frequency converter automatically converts the speech signal into the spectrum of frequencies in real time.

4. The system of claim 1, where the high frequency band comprises a larger range of frequencies than the lower band limited frequency range.

5. The system of claim 1 where the spectral compressor comprises a non-linear compression basis function.

6. The system of claim 1 where the lower band limited frequency range comprises a portion of an analog speech signal bandwidth.

7. The system of claim 1 where the lower band limited frequency range comprises a portion of a telephone bandwidth.

8. The system of claim 1 further comprising a noise detector device that detects and measures a level of noise present when the speech signal is detected.

9. The system of claim 1 further comprising a noise detector device that detects and estimates a level of noise present when the speech signal is detected.

10. The system of claim 1 where the gain controller adjusts the gain of the compressed high frequency band in relation to an independent extraneous signal.

11. The system of claim 1 where the gain controller is coupled to the spectral compressor, and where the gain controller adjusts substantially only the gain of the compressed high frequency band at the lower band limited frequency range.

12. The system of claim 11 where the gain controller applies a plurality of gain adjustments that varies with a signal independent of the detected speech signal.

13. The system of claim 1, where the gain controller amplifies a portion of the speech signal in the compressed high frequency band when the speech signal has a lower signal power level in the compressed high frequency band after compression than before compression.

14. The system of claim 1, where the gain controller attenuates a portion of the speech signal in the compressed high frequency band when the speech signal has a higher signal power level in the compressed high frequency band after compression than before compression.

15. The system of claim 1, where the gain controller selects a level for the variable gain that counteracts an increase or decrease in noise floor in the compressed high frequency band due to the compression of the pre-selected high frequency band into the compressed high frequency band.

16. The system of claim 1, where the gain controller selects a level for the variable gain that substantially aligns the slope of the noise floor present in the compressed high frequency band with the slope of the noise floor present in the uncompressed frequency portion of the speech signal.

17. A speech system that improves the intelligibility of a processed speech, comprising:

a frequency transformer device that converts a speech signal into the frequency domain;

a spectral compressor device coupled to the frequency transformer that compresses a pre-selected high frequency band of the speech signal and maps the compressed high frequency band to a lower frequency band;

a noise detector device that detects and estimates a level of noise present; and

a gain controller device that adjusts a gain of the compressed high frequency band proportionally to a changing level of an independent and extraneous signal, where the gain controller amplifies a portion of the speech signal in the compressed high frequency band when the speech signal has a lower signal power level in the compressed high frequency band after compression than before compression, and where the gain controller attenuates a portion of the speech signal in the compressed high frequency band when the speech signal has a higher signal power level in the compressed high frequency band after compression than before compression;

where the gain controller selects a level for the gain based on a slope of a noise floor present in the compressed high frequency band of the speech signal and a slope of a noise floor present in an uncompressed frequency band of the speech signal.

18. The speech system of claim 17 further comprising a controller that regulates the spectral compressor, the controller comprising a monitor that compares a signal-to-noise ratio of the compressed signal to a signal-to-noise ratio of the signal before it is compressed.

19. The speech system of claim 17 where the gain controller applies a gain that varies with a changing level of the extraneous signal.

20. The speech system of claim 17 where the gain controller applies a variable gain that causes a level of the compressed signal to be substantially coincident with the level of the independent and extraneous signal.

21. The system of claim 17, where the gain controller selects a level for the gain that substantially aligns the slope of the noise floor present in the compressed high frequency band with the slope of the noise floor present in the uncompressed frequency band.

22. A speech system that improves the intelligibility of a processed speech, comprising:

a frequency transformer device that converts a speech signal from time domain into frequency domain in real time;

a spectral compressor device coupled to the frequency transformer that compresses a pre-selected high frequency band of the speech signal and maps the compressed high frequency band to a lower frequency band within a telephone pass band;

a noise detector device that detects and measures a background noise level in the speech signal; and

a gain controller device that applies a variable gain to the compressed high frequency band in relation to the level of the background noise in the speech signal, where the gain controller selects a level for the variable gain that substantially aligns a slope of a noise floor present in the compressed high frequency band with a slope of a noise floor present in an uncompressed frequency portion of the speech signal.

23. The speech system of claim 22 further comprising a controller that regulates the spectral compressor through a communication bus, the controller compares a signal-to-noise ratio of a portion of the detected speech signal to a signal-to-noise ratio of a portion of the compressed signal.

24. The speech system of claim 23 where the controller compares amplitude through a comparison of frequency bins.

25. The speech system of claim 23 further comprising an automatic speech recognition system coupled to the gain controller.

26. The system of claim 22, where the gain controller selects a level for the variable gain that counteracts an increase or decrease in noise floor in the compressed high frequency band due to the compression of the pre-selected high frequency band into the compressed high frequency band.

27. A speech system that improves the intelligibility and quality of a processed speech, comprising:

a frequency transformer device that converts a speech signal into a spectrum of frequencies;

a spectral compressor device electrically coupled to the frequency transformer that compresses a pre-selected high frequency band of the speech signal and maps the compressed high frequency band to a lower band limited frequency range; and

a gain controller device that applies a variable gain to the compressed high frequency band, where the gain controller selects a level for the variable gain that counteracts an increase or decrease in noise floor in the compressed high frequency band due to the compression of the pre-selected high frequency band into the compressed high frequency band, and substantially aligns a slope of the noise floor in the compressed high frequency band with a slope of a noise floor present in an uncompressed frequency portion of the speech signal.