Acoustic signal processing with voice activity detector having processor in an idle state

• US 10,629,226 B1
• Filed: 04/29/2019
• Issued: 04/21/2020
• Est. Priority Date: 10/29/2018
• Status: Active Grant

• Current Assignee: Bestechnic (Shanghai) Co., Ltd.
• Original Assignee: Bestechnic (Shanghai) Co., Ltd.
• Inventors: Tong, Weifeng, Li, Qian, Zhang, Liang
• Primary Examiner(s): Lerner, Martin
• Application Number:
  US16/396,763
• Publication Number:
  US 20200135230A1
• Time in Patent Office: 358 Days
• Field of Search: 704210, 704215, 704216, 704218, 704233, 704251, 704205, 704237
• CPC Class Codes: G10L 15/02 Feature extraction for speech recognition; Selection of recognition unit
• CPC Class Codes: G10L 15/08 Speech classification or search
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/025 Phonemes, fenemes or fenones being the recognition units
• CPC Class Codes: G10L 2015/088 Word spotting
• CPC Class Codes: G10L 2025/783 based on threshold decision
• CPC Class Codes: G10L 25/06 the extracted parameters being correlation coefficients
• CPC Class Codes: G10L 25/18 the extracted parameters being spectral information of each sub-band
• CPC Class Codes: G10L 25/78 Detection of presence or absence of voice signals switching of direction of transmission by voice frequency in two-way loud-speaking telephone systems H04M9/10

Abstract

System and method for acoustic signal processing are disclosed. An exemplary device for acoustic signal processing includes a voice activity detector configured to detect a speech of a user. The device includes a microphone configured to receive acoustic signals from the user. The device further includes at least one processor configured to process the acoustic signals in response to detecting the speech of the user. The at least one processor is in an idle state before the speech of the user is detected.

Claims

1. A device for acoustic signal processing, comprising:

a sensor configured to detect a vibration signal from a user;

a voice activity detector configured to detect a speech of the user; and

a microphone configured to receive an acoustic signal from the user,

wherein the voice activity detector is further configured to detect the speech of the user based on a fusion of a high frequency part of the acoustic signal and a low frequency part of the vibration signal,

wherein the at least one processor is in an idle state before the speech of the user is detected, and

wherein to detect the speech of the user, the voice activity detector is further configured to;

generate a frequency domain representation of the vibration signal based on the vibration signal;

generate a frequency domain representation of the acoustic signal based on the acoustic signal;

compute a correlation coefficient between the frequency domain representation of the vibration signal and the frequency domain representation of the acoustic signal;

detect the speech of the user based on the correlation coefficient; and

determine that the speech of the user is detected when a detected quantity is greater than a threshold, where the detected quantity is expressed by;

2. The device of claim 1, wherein to process the acoustic signal in response to detecting the speech of the user, the at least one processor is further configured to perform keyword detection on the fusion of the acoustic signal and the vibration signal.

3. The device of claim 1, wherein the microphone is a digital microphone or an analog microphone.

4. The device of claim 1, wherein to detect the speech of the user, the voice activity detector is further configured to detect the speech of the user based on the acoustic signal.

5. The device of claim 4, wherein the device further comprises an analog-to-digital converter (ADC) and a filter for processing the acoustic signal.

6. The device of claim 1, further comprising a storage for storing the acoustic signal, wherein the at least one processor is further configured to fetch the acoustic signal from the storage after the speech of the user is detected, wherein the acoustic signal is used for keyword detection and/or speech recognition.

7. The device of claim 1, wherein the correlation coefficient is expressed by:

8. The device of claim 1, further comprising at least one processor configured to process the acoustic signal in response to detecting the speech of the user, and perform detection of keywords on the processed acoustic signal, wherein the keywords are preset as words starting with a vowel with a magnitude of vibration.

9. A system for acoustic signal processing, comprising:

an acoustic signal processing device, comprising;

a sensor configured to detect a vibration signal from a user;

a voice activity detector configured to detect a speech of the user;

a microphone configured to receive an acoustic signal from the user;

at least one processor configured to process the acoustic signal in response to detecting the speech of the user,

wherein the at least one processor is in an idle state before the speech of the user is detected, and

to detect the speech of the user, the voice activity detector is further configured to;

generate a frequency domain representation of the vibration signal based on the vibration signal;

generate a frequency domain representation of the acoustic signal based on the acoustic signal;

compute a correlation coefficient between the frequency domain representation of the vibration signal and the frequency domain representation of the acoustic signal;

detect the speech of the user based on the correlation coefficient; and

calculate a detected quantity in a frequency range based on the correlation coefficient, where the detected quantity is expressed by;

10. The system of claim 9, further comprising a computing device configured to perform keyword detection on the processed acoustic signal.

11. The system of claim 9, wherein the correlation coefficient between the frequency domain representation of the vibration signal and the frequency domain representation of the acoustic signal is computed based on a cross power spectral density of the vibration signal and the acoustic signal.

12. The system of claim 11, wherein the correlation coefficient is expressed by:

13. The system of claim 9, wherein the voice activity detector is further configured to determine that the speech of the user is detected when the detected quantity is greater than a threshold.

14. The system of claim 9, wherein the at least one processor is further configured to perform keyword detection on presetting keywords as words starting with a vowel with a magnitude of vibration.

15. A device for acoustic signal processing, comprising:

a sensor configured to detect a vibration signal from a user;

a voice activity detector configured to detect a speech of the user; and

a microphone configured to receive an acoustic signal from the user,

wherein the voice activity detector is further configured to detect the speech of the user based on a fusion of the processed acoustic signal and the processed vibration signal, and

wherein the at least one processor is in an idle state before the speech of the user is detected,

wherein to detect the speech of the user, the voice activity detector is further configured to;

generate a frequency domain representation of the vibration signal based on the vibration signal;

generate a frequency domain representation of the acoustic signal based on the acoustic signal;

compute a correlation coefficient between the frequency domain representation of the vibration signal and the frequency domain representation of the acoustic signal;

detect the speech of the user based on the correlation coefficient; and

determine that the speech of the user is detected when a detected quantity is greater than a threshold, where the detected quantity is expressed by;

16. The device of claim 15, wherein to process the acoustic signal in response to detecting the speech of the user, the at least one processor is further configured to perform keyword detection on the fusion of the processed acoustic signal and the processed vibration signal.

17. The device of claim 15,

wherein the correlation coefficient is expressed by:

18. The device of claim 15, further comprising at least one processor configured to process the acoustic signal and the vibration signal in response to detecting the speech of the user by filtering the acoustic signal using a high pass filter and by filtering the vibration signal using a low pass filter, and perform detection of keywords on the processed acoustic signal, wherein the keywords are preset as words starting with a vowel with a magnitude of vibration.

19. The device of claim 15, wherein the at least one processor is further configured to perform detection of keywords on the processed acoustic signal, wherein the keywords are preset as words starting with a vowel with a magnitude of vibration.