Periodic signal enhancement system

• US 20060089958A1
• Filed: 10/26/2004
• Published: 04/27/2006
• Est. Priority Date: 10/26/2004
• Status: Active Grant

• Current Assignee: Blackberry Limited
• Original Assignee: Harman Becker Automotive Systems GmbH (Samsung Group)
• Inventors: Giesbrecht, David, Hetherington, Phillip
• Granted Patent: US 7,680,652 B2
• Time in Patent Office: Days
• US Class Current: 708/300
• CPC Class Codes: G10L 21/0364 for improving intelligibility
• CPC Class Codes: G10L 25/90 Pitch determination of speech signals
• CPC Class Codes: H03H 21/0012 Digital adaptive filters
• CPC Class Codes: H03H 2220/08 Variable filter length

Abstract

A signal enhancement system improves the understandability of speech or other audio signals. The system reinforces selected parts of the signal, may attenuate selected parts of the signal, and may increase SNR. The system includes delay logic, an adaptive filter, and signal reinforcement logic. The adaptive filter may track one or more fundamental frequencies in the input signal and outputs a filtered signal. The filtered signal may approximately reproduce the input signal approximately delayed by an integer multiple of the signal'"'"'s fundamental frequencies. The reinforcement logic combines the input signal and the filtered signal output to produce an enhanced signal output.

Claims

1. A signal enhancement system comprising:

a signal input;

delay logic coupled to the signal input;

an adaptive filter coupled to the delay logic; and

reinforcement logic coupled to the adaptive filter and the signal input.

2. The system of claim 1, where the delay logic and adaptive filter implement a glottal pulse delay for voiced sounds.

3. The system of claim 1, where the adaptive filter is a sample-by-sample adaptive filter.

4. The system of claim 1, further comprising gain control logic coupled to the filter.

5. The system of claim 4, where the gain control logic is coupled to the filter and the signal reinforcement logic.

6. The system of claim 1, further comprising an error estimator coupled to the filter and the signal input.

7. The system of claim 6, further comprising gain control logic coupled to the filter and being responsive to the error estimator.

8. The system of claim 1, where the delay logic implements a delay corresponding to a period of a maximum pitch to be enhanced.

9. The system of claim 8, where the maximum pitch comprises a selected maximum voice pitch.

10. The system of claim 8, where the maximum pitch is between approximately 70 Hz and about 400 Hz.

11. The system of claim 8, where the maximum pitch is a female voice maximum pitch.

12. The system of claim 8, where the maximum pitch is a male voice maximum pitch.

13. The system of claim 8, where the filter comprises a length L filter, where L extends the filter over a selected enhancement frequency range.

14. The system of claim 13, where the enhancement frequency range comprises an audible voice frequency band.

15. The system of claim 1, where the signal input is coupled to a slowly adapting signal filter output.

16. The system of claim 1, where the signal input is coupled to a first stage signal filter comprising quasi-stationary frequency tracking and attenuation logic.

17. A signal enhancement system comprising:

a signal input;

means for delaying an input signal received through the signal input;

means for adaptively filtering the delayed signal; and

means for reinforcing the input signal with the filtered output signal.

18. The system of claim 17, further comprising:

means for estimating an error between the filtered output signal and the input signal.

19. The system of claim 17, further comprising:

means for controlling gain of the filtered output signal.

20. The system of claim 18, further comprising:

means for controlling gain of the filtered output signal that is responsive to the means for estimating error.

21. The system of claim 17, further comprising means for filtering a quasi-stationary signal coupled to the signal input.

22. A signal enhancement system comprising:

a signal input;

frequency tracking logic coupled to the signal input and comprising an output, the frequency tracking logic operable to output a fundamental frequency of an input signal;

an adder coupled to the signal input and the output that reinforces the input signal with the fundamental frequency.

23. The signal enhancement system of claim 22, further comprising delay logic coupled to the signal input and the tracking logic, where the tracking logic outputs the fundamental frequency based on a delayed version of the input signal.

24. The signal enhancement system of claim 22, where the fundamental frequency comprises a frequency occurring in a voice frequency range.

25. The signal enhancement system of claim 22, where the fundamental frequency comprises a fundamental frequency created by a male.

26. The signal enhancement system of claim 22, where the fundamental frequency comprises a voice fundamental frequency created by a female.

27. The signal enhancement system of claim 22, further comprising gain control logic coupled to the output.

28. The signal enhancement system of claim 27, further comprising error estimation logic coupled to the signal input and the output, where the gain control logic reduces a gain applied to the tracking output as error increases.

29. The signal enhancement system of claim 27, further comprising a slowly adapting first stage filter coupled to the signal input.

30. The signal enhancement system of claim 29, where the first stage filter comprises quasi-stationary signal tracking and attenuation logic.

31. A method for enhancing a signal, comprising:

receiving an input signal;

delaying the input signal;

estimating periodicity in the input signal based on a delayed signal; and

reinforcing the input signal with the delayed signal.

32. The method of claim 31, where estimating comprises:

adaptively filtering the delayed signal.

33. The method of claim 31, where adaptively filtering comprises:

applying a filter to the delayed signal; and

adapting the filter using Normalized Least Means Squares techniques.

34. The method of claim 31, further comprising:

outputting a filtered output signal; and

determining an error between the filtered output signal and the input signal.

35. The method of claim 34, where estimating comprises:

adapting the filter coefficients based on the error.

36. The method of claim 34, further comprising:

adjusting gain applied to the filtered output signal based on the error.

37. The method of claim 31 where estimating comprises:

estimating a fundamental frequency of the input signal.

38. The method of claim 37, where the fundamental frequency occurs in a voice frequency band.

39. The method of claim 37, where the fundamental frequency is less than approximately 400 Hz.

40. The method of claim 37, where the fundamental frequency is less than approximately 200 Hz.

41. A product comprising:

a machine readable medium; and

machine readable instructions embodied on the machine readable medium that;

delay an input signal;

estimate periodicity in the input signal based on a delayed signal; and

reinforce the input signal with the delayed signal.

42. The product of claim 41, where the instructions for estimating comprise:

adaptive filtering instructions.

43. The product of claim 41, where the instructions for estimating comprise:

adaptive filtering instructions that estimate a fundamental frequency component of the input signal.

44. The method of claim 41, where the instructions for estimating comprise:

Normalized Least Means Squares adaptive filtering instructions.

45. The product of claim 43, where the fundamental frequency occurs in a voice frequency band.

46. The product of claim 42, further comprising instructions that:

dynamically change a filter length.

47. The product of claim 41, further comprising instructions that:

dynamically change an input signal delay.

48. The product of claim 41, further comprising instructions that:

prevent adaptation of an adaptive filter based on an expected input signal characteristic.

49. The product of claim 41, further comprising instructions that:

produce an output signal comprising the frequency content; and

determine an error between the output signal and the input signal.

50. The product of claim 49, further comprising instructions that:

adjust gain applied to the filtered output signal based on the error.

51. The product of claim 41, further comprising instructions that:

track and attenuate a quasi-stationary signal component in the input signal.