System for suppressing passing tire hiss

• US 8,521,521 B2
• Filed: 09/01/2011
• Issued: 08/27/2013
• Est. Priority Date: 05/09/2005
• Status: Active Grant

• Current Assignee: Blackberry Limited
• Original Assignee: QNX Software Systems Limited (Canada) (Blackberry Limited)
• Inventors: Hetherington, Phillip A., Paranjpe, Shreyas A.
• Primary Examiner(s): JACKSON, JAKIEDA R
• Application Number:
  US13/223,863
• Publication Number:
  US 20110311068A1
• Time in Patent Office: 726 Days
• Field of Search: 704/226, 704/225
• US Class Current: 704/226
• CPC Class Codes: G10L 21/0208 Noise filtering

Abstract

A voice enhancement logic improves the perceptual quality of a processed voice. The voice enhancement system includes a passing tire hiss noise detector and a passing tire hiss noise attenuator. The passing tire hiss noise detector detects a passing tire hiss noise by modeling the passing tire hiss. The passing tire hiss noise attenuator dampens the passing tire hiss noise to improve the intelligibility of a speech signal.

Claims

1. A passing tire hiss noise attenuation system, comprising:

a noise detector configured to compare an input signal to a passing tire hiss model and identify whether a noise in the input signal is passing tire hiss; and

a noise attenuator coupled with the noise detector and configured to attenuate at least a portion of the identified passing tire hiss from the input signal to generate an output signal with reduced passing tire hiss noise.

2. The system of claim 1, where the noise detector is configured to identify whether the input signal includes the passing tire hiss by fitting a function to a portion of the input signal.

3. The system of claim 1, where the noise detector is configured to identify whether the input signal includes the passing tire hiss by fitting a function to a portion of the input signal in a time-frequency domain.

4. The system of claim 1, where the noise detector is configured to identify whether the input signal includes the passing tire hiss by fitting a Lorentzian function to a portion of the input signal in a time-frequency domain.

5. The system of claim 1, where the noise detector is configured to identify whether the input signal includes the passing tire hiss by fitting a smoothly varying function to a portion of the input signal.

6. The system of claim 1 where the noise detector is configured to separate noise-like segments of the input signal from remaining portions of the input signal, and where the noise detector is configured to analyze the noise-like segments to identify whether the noise-like segments include passing tire hiss noise.

7. The system of claim 6 where the noise detector is configured to derive the passing tire hiss model when the noise-like segments include passing tire hiss noise, where the noise detector is configured to store the passing tire hiss model in memory, and where the noise attenuator is configured to use the passing tire hiss model stored in memory to remove passing tire hiss from the input signal.

8. The system of claim 1, where the noise detector is configured to receive information from an automotive bus about whether windows of a vehicle are open or closed, and where the noise detector is configured to disable or constrain passing tire hiss noise detection when the information indicates that the windows are closed.

9. The system of claim 1 where the noise detector comprises a processor configured to run logic to detect the passing tire hiss from the input signal.

10. A method of attenuating passing tire hiss noise, comprising:

receiving an input signal;

identifying, by a noise detector that comprises a processor configured to run logic to detect passing tire hiss, whether a noise in the input signal is passing tire hiss based on a comparison between the input signal and a passing tire hiss model; and

attenuating at least a portion of the identified passing tire hiss from the input signal to generate an output signal with reduced passing tire hiss noise.

11. The method of claim 10, where the step of identifying comprises identifying whether the input signal includes the passing tire hiss by fitting a function to a portion of the input signal.

12. The method of claim 10, where the step of identifying comprises identifying whether the input signal includes the passing tire hiss by fitting a function to a portion of the input signal in a time-frequency domain.

13. The method of claim 10, where the step of identifying comprises identifying whether the input signal includes the passing tire hiss by fitting a Lorentzian function to a portion of the input signal in a time-frequency domain.

14. The method of claim 10, where the step of identifying comprises identifying whether the input signal includes the passing tire hiss by fitting a smoothly varying function to a portion of the input signal.

15. The method of claim 10, where the step of identifying comprises:

separating noise-like segments of the input signal from remaining portions of the input signal; and

analyzing the noise-like segments to identify whether the noise-like segments include passing tire hiss noise.

16. The method of claim 15, further comprising:

deriving the passing tire hiss model when the noise-like segments include passing tire hiss noise;

storing the passing tire hiss model in memory; and

removing passing tire hiss from the input signal based on the passing tire hiss model stored in memory.

17. The method of claim 10, further comprising:

receiving information from an automotive bus about whether windows of a vehicle are open or closed; and

disabling or constraining passing tire hiss noise detection when the information indicates that the windows are closed.

18. A non-transitory computer-readable medium with instructions stored thereon, where the instructions are executable by a processor to cause the processor to perform the steps of:

comparing an input signal to a passing tire hiss model;

identifying whether a noise in the input signal is passing tire hiss based on the comparison between the input signal and the passing tire hiss model; and

attenuating at least a portion of the identified passing tire hiss from the input signal to generate an output signal with reduced passing tire hiss noise.

19. The non-transitory computer-readable medium of claim 18, where the step of identifying comprises the step of identifying whether the input signal includes the passing tire hiss by fitting a function to a portion of the input signal in a time-frequency domain.

20. The non-transitory computer-readable medium of claim 18, where the step of identifying comprises identifying whether the input signal includes the passing tire hiss by fitting a smoothly varying function to a portion of the input signal.