Acoustic environment recognizer for optimal speech processing

Patent metadata

Field	Value
Publication number	US-11031027-B2
Application number	US-201815912230-A
Country	US
Kind code	B2
Filing date	Mar 5, 2018
Priority date	Oct 31, 2014
Publication date	Jun 8, 2021
Grant date	Jun 8, 2021

Abstract

A system for providing an acoustic environment recognizer for optimal speech processing is disclosed. In particular, the system may utilize metadata obtained from various acoustic environments to assist in suppressing ambient noise interfering with a desired audio signal. In order to do so, the system may receive an audio stream including an audio signal associated with a user and including ambient noise obtained from an acoustic environment of the user. The system may obtain first metadata associated with the ambient noise, and may determine if the first metadata corresponds to second metadata in a profile for the acoustic environment. If the first metadata corresponds to the second metadata, the system may select a processing scheme for suppressing the ambient noise from the audio stream, and process the audio stream using the processing scheme. Once the audio stream is processed, the system may provide the audio stream to a destination.

First claim

We claim:

1. A system, comprising:
at least one memory that stores a first set of instructions; and
at least one hardware processor that executes the set of instructions to perform a set of operations, the set of operations comprising:
determining whether first metadata associated with ambient noise occurring in an acoustic environment corresponds to second metadata in a profile for the acoustic environment, wherein the second metadata corresponds to a first processing scheme that suppresses the ambient noise from an audio stream including an audio signal associated with a user;
selecting, in response to determining that the first metadata associated with the ambient noise does not correspond to the second metadata in the profile for the acoustic environment, a second processing scheme for a different ambient noise that has a correlation with the ambient noise occurring in the acoustic environment, wherein the second processing scheme is selected based on an estimation of the acoustic environment that is made after the first metadata associated with the ambient noise is determined to not correspond to the second metadata in the profile for the acoustic environment, wherein the second processing scheme suppresses a portion of the ambient noise based on the correlation; and
processing the audio stream using the second processing scheme in order to suppress the portion of the ambient noise in the audio stream interfering with the audio signal, thereby enhancing the audio stream, wherein the second processing scheme is selected prior to any processing of the audio stream by the first processing scheme and the second processing scheme.

2. The system of claim 1 , wherein the set of operations further comprises obtaining, from visual content captured by a camera of a device of the user, the first metadata associated with the ambient noise.

3. The system of claim 1 , wherein the set of operations further comprises providing, after processing the audio stream, the audio stream to a destination.

4. The system of claim 1 , wherein the first set of operations further comprises estimating, if the first metadata associated with the ambient noise occurring in the acoustic environment does not correspond to the second metadata in the profile for the acoustic environment, what the acoustic environment is based on the first metadata.

5. The system of claim 1 , wherein the first set of operations further comprises obtaining, from sensor data, third metadata for refining acoustic parameters utilized in processing the audio stream to suppress the ambient noise in the audio stream.

6. The system of claim 1 , wherein the set of operations further comprises providing the audio stream enhanced by the processing to a speech recognition system.

7. The system of claim 1 , wherein the set of operations further comprises utilizing a noise suppression algorithm, an echo-cancellation algorithm, a speech enhancement algorithm, or a combination thereof, to process the audio stream and to suppress the ambient noise.

8. The system of claim 1 , wherein the set of operations further comprises updating the profile for the acoustic environment based on the first metadata.

9. The system of claim 1 , wherein the set of operations further comprises generating a new profile for the acoustic environment if the first metadata does not correspond to the second metadata.

10. A method, comprising:
determining, by at least one hardware processor, whether first metadata associated with ambient noise occurring in an acoustic environment corresponds to second metadata in a profile for the acoustic environment, wherein the second metadata corresponds to a first processing scheme that suppresses the ambient noise from an audio stream including an audio signal associated with a user;
selecting, by the at least one hardware processor, in response to determining that the first metadata associated with the ambient noise does not correspond to the second metadata in the profile for the acoustic environment, a second processing scheme for a different ambient noise that has a correlation with the ambient noise occurring in the acoustic environment, wherein the second processing scheme is selected based on an estimation of the acoustic environment that is made after the first metadata associated with the ambient noise is determined to not correspond to the second metadata in the profile for the acoustic environment, wherein the second processing scheme suppresses a portion of the ambient noise based on the correlation; and
processing, by the at least one hardware processor, the audio stream using the second processing scheme in order to suppress the portion of the ambient noise in the audio stream interfering with the audio signal, thereby enhancing the audio stream, wherein the second processing scheme is selected prior to any processing of the audio stream by the first processing scheme and the second processing scheme.

11. The method of claim 10 , further comprising providing, after processing the audio stream, the audio stream to a destination.

12. The method of claim 10 , further comprising updating the profile for the acoustic environment based on the first metadata.

13. The method of claim 10 , further comprising obtaining, from sensor data, third metadata for refining acoustic parameters utilized in processing the audio stream to suppress the ambient noise in the audio stream.

14. The method of claim 10 , further comprising utilizing a user profile including a speech characteristic of the user to enhance a quality of the audio stream.

15. The method of claim 10 , further comprising utilizing a noise suppression algorithm, an echo-cancellation algorithm, a speech enhancement algorithm, or a combination thereof, to process the audio stream and to suppress the ambient noise.

16. A computer-readable device comprising a set of instructions which, when executed by at least one hardware processor, cause the at least one hardware processor to perform a set of operations, the set of operations comprising:
determining whether first metadata associated with ambient noise occurring in an acoustic environment corresponds to second metadata in a profile for the acoustic environment, wherein the second metadata corresponds to a first processing scheme that suppresses the ambient noise from an audio stream including an audio signal associated with a user;
selecting, in response to determining that the first metadata associated with the ambient noise does not correspond to the second metadata in the profile for the acoustic environment, a second processing scheme for a different ambient noise that has a correlation with the ambient noise occurring in the acoustic environment, wherein the second processing scheme is selected based on an estimation of the acoustic environment that is made after the first metadata associated with the ambient noise is determined to not correspond to the second metadata in the profile for the acoustic environment, wherein the second processing scheme suppresses a portion of the ambient noise based on the correlation; and
processing the audio stream using the second processing scheme in order to suppress the portion of the ambient noise in the audio stream that is interfering with the audio signal, thereby enhancing the audio stream, wherein the second processing scheme is selected prior to any processing of the audio stream by the first processing scheme and the second processing scheme.

17. The computer-readable device of claim 16 , wherein the set of operations further comprises obtaining, from visual content captured by

Assignees

• At & T Ip I Lp

Inventors

• Schroeter Horst J
• Bowen Donald J
• Dimitriadis Dimitrios B
• Ji Lusheng

Classifications

• G10L25/72

  for transmitting results of analysis · CPC title

• G10L21/0208

  Noise filtering · CPC title

• G10L21/028 Primary

  using properties of sound source · CPC title

• G10L15/20

  Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech (G10L21/02 takes precedence) · CPC title

• G10L21/0216

  characterised by the method used for estimating noise · CPC title