Apparatus and method of detecting target sound

• US 8,213,263 B2
• Filed: 05/20/2009
• Issued: 07/03/2012
• Est. Priority Date: 10/30/2008
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Kim, Kyu-hong, Oh, Kwang-cheol
• Primary Examiner(s): Keith, Jack W
• Assistant Examiner(s): HULKA, JAMES R
• Application Number:
  US12/469,557
• Publication Number:
  US 20100110834A1
• Time in Patent Office: 1,140 Days
• Field of Search: 367/118, 367/124, 367/125, 181/125
• US Class Current: 367/125
• CPC Class Codes: G01S 3/8083 determining direction of source

Abstract

An apparatus and method of voice activity detection (VAD) are disclosed. To detect a target sound, a target sound detecting apparatus calculates the phase difference corresponding to each frequency component from a plurality of sound signals converted in frequency domain, calculates a characteristic value indicating possibility that the phase difference of the frequency component is within an allowable phase difference range of target sound calculated based on a direction angle of the target sound, and detects the presence and/or absence of target sound using the characteristic value.

Claims

1. An apparatus configured to detect a target sound based on each sound signal acquired by a plurality of acoustic sensors, comprising:

a frequency converter configured to convert the sound signal into a frequency input signal;

a phase difference calculator configured to calculate a phase difference corresponding to each frequency component using the frequency input signal;

a characteristic value extractor configured to extract a characteristic value indicating a possibility that the phase difference corresponding to the frequency component is within an allowable phase difference range of target sound determined based on a direction angle of the target sound; and

a determining unit configured to compare the characteristic value to a threshold characteristic value, so as to determine whether the target sound exists in the sound signal according to the result of the comparison.

2. The apparatus of claim 1, wherein the frequency converter segments the sound signal into a plurality of frames, and converts the plurality of frames into the frequency input signal in units of frames.

3. The apparatus of claim 1, wherein the characteristic value extractor calculates the allowable phase difference range of target sound based on an allowable angle range of target sound including the direction angle of target sound.

4. The apparatus of claim 3, wherein the direction angle of the target sound and the allowable angle range of the target sound including the direction angle of the target sound are adjustable.

5. The apparatus of claim 1, wherein the characteristic value is calculated using the number of phase differences of frequency components included in the allowable phase difference range of the target sound.

6. The apparatus of claim 5, wherein the characteristic value is an average number of valid frequency components obtained by dividing the number of frequency components included in the allowable phase difference range of the target sound by a total number of frequency components of the frequency input signal.

7. The apparatus of claim 1, further comprising a target sound extractor configured to extract the target sound from the sound signal using the result of the determination.

8. The apparatus of claim 1, wherein the sound signal has a magnitude of sound pressure and a phase of sound satisfying a far-field condition.

9. The apparatus of claim 1, wherein the allowable phase difference range of target sound is bounded by a lowest threshold value and a highest threshold value, the lowest and highest threshold values respectively determined by the following equations:

10. A method of detecting target sound based on each sound signal acquired by a plurality of acoustic sensors, the method comprising:

converting the sound signal into an frequency input signal;

calculating a phase difference corresponding to each frequency component of the frequency input signal;

obtaining an allowable phase difference range of target sound, based on a direction angle of the target sound and an allowable angle range of the target sound including the direction angle of the target sound;

extracting a characteristic value indicating a possibility that the phase difference corresponding to the frequency component is within an allowable phase difference range of the target sound; and

comparing the characteristic value to a threshold characteristic value, so as to determine whether the target sound exists in the sound signal according to the result of the comparison.

11. The method of claim 10, wherein the converting of the sound signal into the frequency input signal comprises:

segmenting the sound signal into a plurality of frames; and

converting the plurality of frames into the frequency input signal in units of frames.

12. The method of claim 10, wherein the characteristic value is calculated using the number of phase difference of frequency components included in the allowable phase difference range of the target sound.

13. The method of claim 12, wherein the characteristic value is an average number of valid frequency components obtained by dividing the number of frequency components included in the allowable phase difference range of the target sound by a total number of frequency components of the frequency input signal.

14. The method of claim 10, further comprising:

extracting the target sound from the sound signal, based on the result of the determination.

15. The method of claim 10, wherein the direction angle of the target sound and the allowable angle range of the target sound including the direction angle of target sound are adjustable.

16. The method of claim 10, wherein the sound signal satisfies a far-field condition that the sound source is far away from the plurality of sensors.

17. The method of claim 10, wherein the allowable phase difference range of target sound is bounded by a lowest threshold value and a highest threshold value, the lowest and highest threshold values respectively determined by the following equations:

18. A non-transitory computer-readable storage medium configured to store a program to detect a target sound, comprising instructions to cause a computer to:

convert the sound signal into a frequency input signal;

calculate a phase difference corresponding to each frequency component of the frequency input signal;

obtain an allowable phase difference range of target sound, based on a direction angle of the target sound and an allowable angle range of the target sound including the direction angle of the target sound;

extract a characteristic value indicating a possibility that the phase difference corresponding to the frequency component is within an allowable phase difference range of the target sound; and

compare the characteristic value to a threshold characteristic value, so as to determine whether the target sound exists in the sound signal according to the result of the comparison.

19. The computer-readable storage medium of claim 18, wherein the allowable phase difference range of target sound is bounded by a lowest threshold value and a highest threshold value, the lowest and highest threshold values respectively determined by the following equations: