Systems and methods for identifying a sound event

• US 9,812,152 B2
• Filed: 02/06/2015
• Issued: 11/07/2017
• Est. Priority Date: 02/06/2014
• Status: Active Grant

• Current Assignee: Analog Devices, Inc.
• Original Assignee: OtoSense, Inc. (Analog Devices, Inc.)
• Inventors: Christian, Sebastien J. V.
• Primary Examiner(s): Flanders, Andrew C
• Application Number:
  US14/616,627
• Publication Number:
  US 20150221321A1
• Time in Patent Office: 1,005 Days
• CPC Class Codes: G08B 1/08 using electric transmission ; transformation of alarm signals to electrical signals from a different medium, e.g. transmission of an electric alarm signal upon detection of an audible alarm signal
• CPC Class Codes: G08B 17/10 Actuation by presence of smoke or gases , e.g. automatic alarm devices for analysing flowing fluid materials by the use of optical means
• CPC Class Codes: G08B 21/18 Status alarms G08B21/02 takes precedence
• CPC Class Codes: G10H 2210/301 Soundscape or sound field simulation, reproduction or control for musical purposes, e.g. surround or 3D sound; Granular synthesis
• CPC Class Codes: G10L 21/14 by displaying frequency domain information
• CPC Class Codes: G10L 25/27 characterised by the analysis technique
• CPC Class Codes: G10L 25/48 specially adapted for particular use

Abstract

Systems and methods for identifying a perceived sound event are provided. In one exemplary embodiment, the system includes an audio signal receiver, a processor, and an analyzer. The system deconstructs a received audio signal into a plurality of audio chunks, for which one or more sound identification characteristics are determined. One ore more distances of a distance vector are then calculated based on one or more of the sound identification characteristics. The distance vector can be a sound gene that serves as an identifier for the sound event. The distance vector for a received audio signal is compared to distance vectors of predefined sound events to identify the source of the received audio signal. A variety of other systems and methods related to sound identification are also provided.

Claims

1. A method for identifying sound events, comprising:

receiving one or more signals corresponding to incoming sound events;

for each of one or more of the incoming sound events;

deconstructing the corresponding signal into one or more audio chunks;

determining one or more sound identification characteristics based on the corresponding one or more audio chunks;

generating a sound vector based on the corresponding sound identification characteristics;

determining, in real time, if the incoming sound event matches one or more of a plurality of predefined sound events, the determination being performed by each of the predefined sound events for its respective predefined sound event; and

identifying the incoming sound event based on the determination performed by the plurality of predefined sound events.

2. The method of claim 1,

wherein the determining if the incoming sound event matches one or more of the plurality of predefined sound events comprises:

comparing the sound vector of the incoming sound event to the sound vectors of the plurality of predefined sound events; and

generating, based on the comparison, a distance vector comprising a calculated distance between the one or more sound identification characteristics of the incoming sound event and corresponding sound identification characteristics of each of the plurality of predefined sound events included in the sound vectors of the plurality of predefined sound events.

3. The method of claim 1, wherein the one or more calculated distances of the distance vector include one or more distances that are representative of at least one of:

a Soft Surface Change History, Main Ray Histories Matching, Surface Change History Autocorrelation Matching, Spectral Evolution Signature Matching, a Pulse Number Comparison, a location, a time, a day, a position of a device that receives the signal from the incoming sound event, an acceleration of the device that receives the signal from the incoming sound event, and a light intensity detected by the device that receives the signal of the incoming sound event.

4. The method of claim 1, wherein the adjusting of the one or more commensurate values in the sound vectors of the plurality of predefined sound events according to the user-provided information further comprises adjusting relative weights of one or more dimensions of the sound vectors of one or more of the plurality of predefined sound events.

5. The method of claim 1, further comprising, prior to or during the step of comparing in the sound vector of the incoming sound event to the sound vectors of the plurality of predefined sound events stored in a database, optimizing the comparing step by eliminating from consideration one or more of the plurality of predefined sound events based on commensurate information relating to the incoming sound event and the one or more of the plurality of predefined sound events.

6. The method of claim 5, wherein the optimizing the comparing step further comprises performing at least one of the following optimization steps:

performing a Strong Context Filter, performing a Scan Process, and performing a Primary Decision Module.

7. The method of claim 1, further comprising:

identifying which of the one or more the sound identification characteristics of the sound vector of the incoming sound event or of the plurality of predefined sound events have the greatest impact on the identity of the incoming sound event; and

comparing at least a portion of the one or more identified sound identification characteristics of the sound vector of the incoming sound event to the commensurate sound identification characteristics of a sound vector of the plurality of predefined sound events before comparing other sound identification characteristics of the sound vector of the incoming sound event to the other commensurate sound identification characteristics of the sound vector of the plurality of predefined sound events.

8. The method of claim 1, further comprising:

prior to the determining the one or more sound identification characteristics of the incoming sound event based on the corresponding one or more audio chunks;

multiplying the one or more audio chunks, by a Hann window; and

performing a Discrete Fourier Transform on the one or more audio chunk that is multiplied by the Hann window.

9. The method of claim 8, further comprising:

performing a logarithmic ratio on the one or more audio chunks after the Discrete Fourier Transform is performed; and

rescaling a result after the logarithmic ratio is performed.

10. The method of claim 1, wherein the one or more sound identification characteristics include at least one of:

a Soft Surface Change History, a Soft Spectrum Evolution History, a Spectral Evolution Signature, a Main Ray History, a Surface Change Autocorrelation, a Pulse Number, a location, a time, a day, a position of a device that receives the signal from the incoming sound event, an acceleration of the device that receives the signal from the incoming sound event, and a light intensity detected by the device that receives the signal of incoming sound event.

11. The method of claim 1, wherein the determining, in real time, if the incoming sound event matches one or more of the plurality of predefined sound events is performed at least partially simultaneously by two or more of the plurality of predefined sound events.

12. The method of claim 1, wherein the determining, in real time, if the incoming sound event matches one or more of the plurality of predefined sound events is performed at least partially simultaneously by two or more of the plurality of predefined sound events for two or more of the incoming sound events.

13. The method of claim 1, further comprising:

outputting the identity of the incoming sound event;

receiving user-provided information relating to the incoming sound event; and

adjusting one or more commensurate values in sound vectors of the plurality of predefined sound events based on the user-provided information.

14. A method for creating a sound identification gene, comprising:

deconstructing one or more audio signals into a plurality of audio chunks;

determining one or more sound identification characteristics for one or more audio chunks of the plurality of audio chunks;

calculating one or more values of sound vectors for the one more audio signals based on the corresponding one or more sound identification characteristics; and

formulating sound identification genes corresponding to the one or more audio signals based on an N-dimensional comparison of the calculated one or more values of the sound vectors with one or more values of sound vectors of predefined sound events stored in a database, where N represents the number of calculated values,

wherein the N-dimensional comparison is performed by the predefined sound events.

15. The method of claim 14, further comprising adjusting a profile of the sound identification genes for the one or more of the one or more predefined sound events according to the user-provided information by adjusting relative weights of one or more values of the sound vectors for the sound identification genes.

16. The method of claim 15, wherein the adjusting of the relative weights further comprises adjusting a hyper-plane extending between identified true positive results and identified false positive results for the sound identification genes.

17. The method of claim 14, wherein the one or more sound identification characteristics include at least one of:

a Soft Surface Change History, a Soft Spectrum Evolution History, a Spectral Evolution Signature, a Main Ray History, a Surface Change Autocorrelation, a Pulse Number, a location, a time, a day, a position of a device that receives the one or more audio signals, an acceleration of the device that receives the one more audio signals, and a light intensity detected by the device that receives the one or more audio signals.

18. The method of claim 14, wherein the one or more values of the sound vectors include one or more distances that are representative of at least one of:

a Soft Surface Change History, Main Ray Histories Matching, Surface Change History Autocorrelation Matching, Spectral Evolution Signature Matching, a Pulse Number Comparison, a location, a time, a day, a position of a device that receives the one or more audio signals, an acceleration of the device that receives the one or more audio signals, and a light intensity detected by the device that receives the one or more audio signals.

19. The method of claim 14, wherein the N-dimensional comparison is performed at least partially simultaneously by two or more of the predefined sound events.

20. The method of claim 19, wherein the N-dimensional comparison is performed at least partially simultaneously by two or more of the predefined sound events for two or more of the audio signals.

21. The method of claim 14, further comprising:

outputting the sound identification genes of the one or more audio signals;

receiving user-provided information related to the one or more audio signals; and

adjusting a profile of the sound identification genes of one or more of the one or more predefined sound events according to the user-provided information.

22. A method for identifying a sound event, comprising:

receiving, via an audio signal receiver of a sound identification system, a signal from an incoming sound event;

deconstructing, by a processor of the sound source identification system, the signal into a plurality of audio chunks;

determining, by the processor, one or more sound identification characteristics of the incoming sound event for one or more audio chunks of the plurality of audio chunks;

calculating, by the processor, one or more values of a sound vector of the incoming sound for each of the one or more sound identification characteristics;

identifying, by the processor, which of the one or more values of the sound vector of the incoming sound event or one or more predefined sound events have the greatest impact on determining the identity of the incoming sound event;

comparing, by the processor, in real time the sound vector of the incoming sound event to a sound vector of the one or more predefined sound events stored in a database and calculating, by the processor, one or more commensurate distances of a distance vector for each of the one or more sound identification characteristics with respect to each of the one or more predefined sound events, wherein one or more of the identified values of the sound vector of the incoming sound event having the greatest impact on determining the identity of the incoming sound event are compared to the commensurate values of the sound vector of the one or more predefined sound events before other values of the sound vector of the incoming sound event are compared to the other commensurate values of the sound vector of the one or more predefined sound events;

identifying, by the processor, the incoming sound event based on the comparison of the one or more commensurate distances of the distance vector between the incoming sound event and each of the one or more predefined sound events stored in the database; and

communicating, by the processor, an identity of the incoming sound event to a user.

23. The method of claim 22, further comprising:

prior to determining one or more sound identification characteristics of the incoming sound event for an audio chunk, multiplying the audio chunk by a Hann window; and

performing a Discrete Fourier Transform on the audio chunk that is multiplied by a Hann window.

24. The method of claim 23, further comprising:

performing a logarithmic ratio on the audio chunk after the Discrete Fourier Transform is performed; and

rescaling a result after the logarithmic ratio is performed.

25. The method of claim 22, wherein the one or more sound identification characteristics include at least one of:

a Soft Surface Change History, a Soft Spectrum Evolution History, a Spectral Evolution Signature, a Main Ray History, a Surface Change Autocorrelation, a Pulse Number, a location, a time, a day, a position of a devices that receives the signal from the incoming sound event, an acceleration of the device that receives the signal from the incoming sound event, and a light intensity detected by the devices that receives the signals of the incoming sound event.

26. The method of claim 22, wherein the one or more distances of a distance vector include one or more distances that are representative of at least one of:

a Soft Surface Change History, Main Ray Histories Matching, Surface Change History Autocorrelation Matching, Spectral Evolution Signature Matching, and a Pulse Number Comparison, a location, a time, a day, a position of a device that receives the signal from the incoming sound event, an acceleration of the device that receives the signal from the incoming sound event, and a light intensity detected by the device that receives the signal of incoming sound event.

27. The method of claim 22, wherein adjusting one or more commensurate values in the sound vector of the predefined sound events stored in the database according to the information related to the incoming sound event received from the user further comprises adjusting relative weights of one or more dimensions of the sound vector for one or more predefined sound events.

28. The method of claim 22, further comprising, prior to or during the step of comparing in real time the sound vector of the incoming sound event to a sound vector of one or more predefined sound events stored in a database, optimizing the comparing step by eliminating from consideration one or more of the predefined sound events based on commensurate information known about the incoming sound event and the one or more predefined sound events.

29. The method of claim 28, wherein optimizing the comparing step further comprises performing at least one of the following optimization steps:

performing a Strong Context Filter, performing a Scan Process, and performing a Primary Decision Module.

30. A sound source identification system, comprising:

an audio signal receiver;

a processor dividing an audio signal received by the audio signal receiver into a plurality of audio chunks, the processor being operable to control;

an analyzer operable to;

determine one or more sound identification characteristics of one or more audio chunks of the plurality of audio chunks,

compare in real time the received audio signal to one or more predefined sound events stored in a database,

calculate one or more distances of a distance vector between the received audio signal and each of the one or more predefined sound events in the database, and

identify the received audio signal based on the distances of calculated distance vectors;

a user interface operable to communicate an identity of the received audio signal to a user; and

an adaptive learning module operable to identify one or more values associated with one or more of the sound identification characteristics of a received audio signal or a predefined sound event that has the greatest impact on determining the identity of the received audio signal so that the identified greatest impact values of the received audio signal and the one or more predefined sound events can be compared before to comparing other values of the received audio signal and the one or more predefined sound events when identifying the received audio signal.

31. The system of claim 30, wherein the one or more sound identification characteristic determined by the analyzer include at least one of:

a Soft Surface Change History, a Soft Spectrum Evolution History, a Spectral Evolution Signature, a Main Ray History, a Surface Change Autocorrelation, and a Pulse Number.

32. The system of claim 30, wherein the one or more distances calculated by the analyzer include one or more distances that are representative of at least one of:

a Soft Surface Change History, Main Ray Histories Matching, Surface Change History Autocorrelation Matching, Spectral Evolution Signature Matching, and a Pulse Number Comparison.

33. The system of claim 30, wherein the user interface is in communication with the analyzer and configured to allow a user to input information that the analyzer can use to adjust at least one of one or more characteristics and one or more distances of the one or more predefined sound events stored in the database.

34. A system, comprising:

at least one memory; and

a processor operable to;

receive, via an audio signal receiver, one or more audio signals;

divide the one or more audio signals into a plurality of audio chunks;

cause an analyzer to;

determine one or more sound identification characteristics based on the plurality of audio chunks;

compare, in real time, the sound identification characteristics of the one or more audio signals to corresponding sound identification characteristics of one or more predefined sound events stored in a database;

calculate, based on the comparison, one or more distances of a distance vector comprising distances between the one or more audio signals and the one or more predefined sound events; and

identify the one or more audio signals based on the calculated distances of the distance vectors,

wherein the comparing and the calculating are performed by each of the one or more predefined sound events.

35. The system of claim 34, wherein the comparing and the calculating are performed at least partially simultaneously by two or more of the one or more predefined sound events.

36. The system of claim 34, wherein the comparing and the calculating are performed at least partially simultaneously by two or more of the plurality of predefined sound events for two or more of the audio signals.

37. The system of claim 34, the processor being operable to:

communicate, via a user interface, an identity of the one or more audio signals to a user;

receive, via the user interface, user-provided information related to the one or more audio signals; and

adjust at least a portion of the one or more sound identification characteristics of one or more of the one or more predefined sound events based on the information received from the user.

38. The system of claim 34, wherein the one or more sound identification characteristics include at least one of:

a Soft Surface Change History, a Soft Spectrum Evolution History, a Spectral Evolution Signature, a Main Ray History, a Surface Change Autocorrelation, and a Pulse Number.

39. The system of claim 34, wherein the one or more distances include one or more distances that are representative of at least one of:

a Soft Surface Change History, Main Ray Histories Matching, Surface Change History Autocorrelation Matching, Spectral Evolution Signature Matching, and a Pulse Number Comparison.

40. The system of claim 34, wherein the processor is further operable to receive user input information for adjusting at least one of the one or more characteristics and the one or more distances corresponding to the one or more predefined sound events.

41. The system of claim 34, wherein the processor is further operable to:

identify which of the one or more of sound identification characteristics of one of the one or more audio signals or of one of the one or more predefined sound events has the greatest impact on determining the identity of the one of the one or more audio signals such that the identified greatest impact sound identification characteristics can be used for the comparing before other sound identification characteristics different than the greatest impact sound identification characteristics.