Patent US10492008

Hearing device with neural network-based microphone signal processing
[+] Show Details

US10492008

1. A hearing system, comprising: (2) (4)

a plurality of microphones configured to receive input sounds including a target sound and to produce a plurality of microphone signals including the target sound;

a control circuit including a neural network and configured to produce an output signal using a processed signal produced by the neural network, the output signal and the processed signal each including the target sound, the neural network configured to receive the plurality of microphone signals and to control directionality of the plurality of microphones by processing the plurality of microphone signals using a nonlinear signal processing algorithm to produce the processed signal; and

a receiver configured to produce an output sound using the output signal,

wherein the neural network includes: an input to receive the plurality of microphone signals;

an output;

a linear signal processing path between the input and the output, the linear signal processing path including one or more linear hidden layers; and

a nonlinear signal processing path between the input and the output, the nonlinear signal processing path including one or more nonlinear hidden layers.

2. The system of claim 1, wherein the neural network is trained to process the plurality of microphone signals for a desirable balance between a signal-to-noise ratio (SNR) and distortion of the target sound. (2) (0)

18. The system of claim 1, comprising a hearing device including the plurality of microphones, the control circuit, and the receiver. (1) (0)

7. A hearing system configured to be worn by a wearer, comprising: (3) (5)

a pair of left and right hearing aids configured to be worn by the wearer and communicatively coupled to each other, the left and right hearing aids each including:

a microphone configured to receive input sounds including a target sound and produce a microphone signal including the target sound;

a control circuit including a neural network and configured to produce an output signal using a processed signal produced by the neural network, the output signal and the processed signal each including the target sound, the neural network configured to receive the microphone signals produced by microphones of the left and right hearing aids and to control directionality of the microphones of the left and right hearing aids using a nonlinear signal processing algorithm to produce the processed signal; and

a receiver configured to produce an output sound using the output signal,

wherein the neural network comprises: an input to receive the microphone signals;

an output;

a linear signal processing path between the input and the output; and

a nonlinear signal processing path between the input and the output.

8. The system of claim 7, further comprising a remote microphone communicatively coupled to the pair of left and right hearing aids and configured to receive a remote sound and produce a remote microphone signal, wherein the control circuit is configured to produce the output signal using the microphone signals produced by microphones of the left and right hearing aids and the remote microphone signal. (0)

9. The system of claim 7, wherein the target sound is a speech, and the neural network is trained to process the microphone signals for a desirable balance between a signal-to-noise ratio (SNR) and distortion of the speech. (0)

10. The system of claim 7, wherein the neural network comprises: (0) (3)

11. A method for operating a hearing system, comprising: (5) (4)

receiving microphone signals including a target sound from a plurality of microphones of the hearing system;

processing the microphone signals using a neural network configured to receive the microphone signals and to produce a processed signal including the target sound using the received microphone signals and a nonlinear signal processing algorithm, including processing the microphone signals using a linear signal processing path and a nonlinear signal processing path, the processing controlling a directionality of the plurality of microphones;

producing an output signal including the target sound using the processed signal; and

producing an output sound based on the output signal using a receiver of the hearing system.

12. The method of claim 11, wherein processing the microphone signals comprises processing the microphone signals using a neural network trained for a cost function and a desirable balance between a signal-to-noise ratio (SNR) and distortion of the target sound. (1) (0)

15. The method of claim 11, wherein receiving the microphone signals comprises receiving a left microphone signal from a left hearing aid and a right microphone signal from a right hearing aid, the left and right hearing aid each being worn in or about an ear of a hearing aid wearer. (0)

16. The method of claim 11, wherein receiving the microphone signals comprises receiving one or more microphone signal from a microphone external to a hearing aid and communicatively coupled to the hearing aid, and processing the microphone signals comprises processing the microphone signals using a processor of the hearing aid. (0)

17. The method of claim 11, wherein processing the microphone signals comprises processing the microphone signals using the processor of the hearing aid and a processor in an external device communicatively coupled to the hearing aid. (0)

20. The method of claim 11, wherein processing the microphone signals using the neural network comprises processing the microphone signals to improve intelligibility of speech in a noisy background. (0)

View Abstract and Specification Size

PDF with Images and Document Face >

Full Text Publication >

Patent US10492008

Hearing device with neural network-based microphone signal processing

Patent Matrix® Search