Neural network-driven feedback cancellation

Abstract

Disclosed herein, among other things, are apparatus and methods for neural network-driven feedback cancellation for hearing assistance devices. Various embodiments include a method of signal processing an input signal in a hearing assistance device to mitigate entrainment, the hearing assistance device including a receiver and a microphone. The method includes performing neural network processing to identify acoustic features in a plurality of audio signals and predict target outputs for the plurality of audio signals, and using the trained neural network to control acoustic feedback cancellation of the input signal.

1 . A method of signal processing an input signal in a hearing device including a receiver and a microphone, the method comprising:
 training a neural network to identify acoustic features in a plurality of audio signals and predict target outputs for the plurality of audio signals; and   using the trained network to control acoustic feedback cancellation on the input signal, including using the neural network to pre-process signals for adaptation of a feedback canceller to mitigate entrainment.   
     
     
         2 . The method of  claim 1 , wherein training the neural network to identify acoustic features in a plurality of audio signals and predict target outputs for the plurality of audio signals includes performing training offline from data collected during normal use of the hearing device. 
     
     
         3 . The method of  claim 1 , wherein the training is performed on an external device. 
     
     
         4 . The method of  claim 3 , wherein the training is performed based on data collected from wearers stored on a server connected to the hearing device by a communication network. 
     
     
         5 . The method of  claim 4 , wherein neural network processing runs on the server and updates parameters of feedback cancellation on the hearing device. 
     
     
         6 . The method of  claim 3 , wherein the training is performed on a mobile device. 
     
     
         7 . The method of  claim 6 , wherein neural network processing runs on the mobile device and updates parameters of feedback cancellation on the hearing device. 
     
     
         8 . The method of  claim 1 , wherein using the trained network to control acoustic feedback cancellation on the input signal includes using the trained network to control subband acoustic feedback cancellation of the input signal. 
     
     
         9 . The method of  claim 1 , comprising training the network to manipulate parameters of adaptive feedback cancellation. 
     
     
         10 . The method of  claim 9 , comprising training the network to select optimal values for parameters that control a rate at which a feedback canceller adapts. 
     
     
         11 . The method of  claim 9 , comprising training the network to control depth or rate of phase modulation. 
     
     
         12 . The method of  claim 9 , comprising training the network to control an adaptation gradient for adaptive feedback cancellation of the input signal. 
     
     
         13 . The method of  claim 1 , comprising training the network to predict or control adaptive feedback cancellation filter coefficients. 
     
     
         14 . The method of  claim 1 , comprising training the network to produce an estimated feedback signal. 
     
     
         15 . The method of  claim 1 , comprising training the network to produce an estimated feedback-free input signal. 
     
     
         16 . The method of  claim 1 , wherein using the trained network to control acoustic feedback cancellation of the input signal includes using the trained network to control acoustic feedback cancellation of the input signal during conditions in which the input signal is a tonal or pitched signal. 
     
     
         17 . A hearing device, comprising:
 a microphone configured to receive audio signals; and   a processor configured to process the audio signals to correct for a hearing impairment of a wearer, the processor further configured to:   train a neural network processing to identify acoustic features in a plurality of audio signals and predict target outputs for the plurality of audio signals; and   control acoustic feedback cancellation on the input signal using the results of neural network processing, including using the neural network to pre-process signals for adaptation of a feedback canceller to mitigate entrainment.   
     
     
         18 . The hearing device of  claim 17 , wherein the hearing device is a completely-in-the-canal (CIC) hearing aid. 
     
     
         19 . The hearing device of  claim 17 , wherein the hearing device is a receiver-in-canal (RFC) hearing aid. 
     
     
         20 . The hearing device of  claim 18 , further comprising multiple microphones configured to receive audio signals.