A hearing device, e.g. a hearing aid or a headset, configured to be worn by a comprises an input unit for providing at least one electric input signal in a time-frequency representation; and a signal processor comprising a target signal estimator for providing an estimate of the target signal; a noise estimator for providing an estimate of the noise; and a gain estimator for providing respective gain values in dependence of said target signal estimate and said noise estimate. The gain estimator comprises a trained neural network, wherein the outputs of the neural network comprise real or complex valued gains, or separate real valued gains and real valued phases. The signal processor is configured—at a given time instance t—to calculate changes Δx(i,t)=x(i,t)−{circumflex over (x)}(i,t−1), and Δh(j,t−1)=h(j,t−1)−ĥ(j,t−2) to an input vector x(t) and to the hidden state vector h(t−1), respectively, from one time instance, t−1, to the next, t, and where {circumflex over (x)}(i,t−1) and ĥ(j,t−2) are estimated values of x(i,t−1) and h(j,t−2), respectively, where indices i, j refers to the i.sup.th input neuron and the j.sup.th neuron of the hidden state, respectively, where 1≤i≤N.sub.ch,x and 1≤j≤N.sub.ch,oh, wherein N.sub.ch,x and N.sub.ch,oh is the number of processing channels of the input vector x and the hidden state vector h, respectively, and wherein the signal processor is further configured to provide that the number of updated channels among said N.sub.ch,x and said N.sub.ch,oh processing channels of the modified gated recurrent unit for said input vector x(t) and said hidden state vector h(t−1), respectively, at said given time instance t is limited to a number of peak values N.sub.p,x, and N.sub.p,oh, respectively, where N.sub.p,x is smaller than N.sub.ch,x, and N.sub.p,oh, is smaller than N.sub.ch,oh.

A hearing device comprises an input transducer comprising a microphone for providing an electric input signal representative of sound in the environment of the hearing device, a pre-processor for processing electric input signal and providing a multitude of feature vectors, each being representative of a time segment thereof, a neural network processor adapted to implement a neural network for implementing a detector configured to provide an output indicative of a characteristic property of the at least one electric input signal, the neural network being configured to receive said multitude of feature vectors as input vectors and to provide corresponding output vectors representative of said output of said detector in dependence of said input vectors. The hearing device further comprises a transceiver comprising a transmitter and a receiver for establishing a communication link to another part or device or server, at least in a particular adaptation-mode of operation, and a selector for—in said particular adaptation-mode of operation—routing said feature vectors to said transmitter for transmission to said another part or device or server, and—in a normal mode of operation—to route said feature vectors to said neural network processor for use as inputs to said neural network, a neural network controller connected to said neural network processor for—in said particular adaptation-mode of operation—receiving optimized node parameters, and to apply said optimized node parameters to said nodes of said neural network to thereby implement an optimized neural network in said neural network processor, wherein the optimized node parameters have been selected among a multitude of sets of node parameters for respective candidate neural networks according to a predefined criterion in dependence of said feature vectors. A method of selecting optimized parameters for a neural network for use in a portable hearing device is further disclosed. The invention may e.g. be used in hearing aids or headsets, or similar, e.g. wearable, devices.

A hearing device comprises a) at least one input transducer configured to pick up sound from an acoustic environment around the user when the user is wearing the hearing device, the at least one input transducer providing at least one electric input signal representative of said sound, b) at least one analysis filter bank configured to provide said at least one electric input signal as a multitude of frequency sub-band signals, the at least one analysis filter bank comprising b1) a plurality of M first filters h.sub.m(n), whose impulse responses are modulated from a first prototype filter h(n), where m=0, 1, . . . , M−1 is a frequency band index, and n is a time index, c) a processor for processing said at least one electric input signal provided by said at least one analysis filter bank, or a signal originating therefrom, and providing a processed signal, d) an output transducer configured to provide stimuli perceivable as sound to the user in dependence of said processed signal, and e) a controller for controlling said analysis filter bank by applying a different first prototype filter to said at least one filter bank in dependence of said current acoustic environment. A method of operating a hearing device is further disclosed.

A method for detecting a prominent tone of an input audio includes establishing a first analysis audio signal based on the input audio signal, establishing a second analysis audio signal based on the input audio signal, wherein an analysis audio signal of the first analysis audio signal and the second analysis audio signal is established by applying an analysis audio filter to the input audio signal, comparing the first analysis audio signal and the second analysis audio signal to obtain an energy level contrast, and determining a representation of the prominent tone by converting the energy level contrast by a contrast-to-frequency mapping function.

Disclosed is a personal hearing device, an external acoustic processing device and an associated computer program product. The personal hearing device includes: a microphone, for receiving an input acoustic signal, wherein the input acoustic signal is a mixture of sounds coming from a first acoustic source and from other acoustic source(s); a speaker; and an acoustic processing circuit, for automatically distinguishing within the input acoustic signal the sound of the first acoustic source from the sound of other acoustic source(s); wherein the acoustic processing circuit further processes the input acoustic signal by having different modifications to the sound of the first acoustic source and to the sound of other acoustic source(s), whereby the acoustic processing circuit produces an output acoustic signal to be played on the speaker.

A hearing device, e.g. a hearing aid, configured to be worn by a user, comprises a) two or more input transducers (e.g. microphones) wherein said two or more input transducers during use of the hearing device are arranged with a distance between them; b) a directional system comprising a directional algorithm configured to provide a directional pattern in dependence of said distance. The hearing device is configured to estimate a current value of said distance, or an equivalent acoustic delay, or beamformer weights of said directional system, thereby the directional performance can be optimized to the individual user.

A hearing device comprises an input transducer comprising a microphone for providing an electric input signal representative of sound in the environment of the hearing device, a pre-processor for processing electric input signal and providing a multitude of feature vectors, each being representative of a time segment thereof, a neural network processor adapted to implement a neural network for implementing a detector configured to provide an output indicative of a characteristic property of the at least one electric input signal, the neural network being configured to receive said multitude of feature vectors as input vectors and to provide corresponding output vectors representative of said output of said detector in dependence of said input vectors. The hearing device further comprises a transceiver comprising a transmitter and a receiver for establishing a communication link to another part or device or server, at least in a particular adaptation-mode of operation, and a selector for—in said particular adaptation—mode of operation—routing said feature vectors to said transmitter for transmission to said another part or device or server, and—in a normal mode of operation—to route said feature vectors to said neural network processor for use as inputs to said neural network, a neural network controller connected to said neural network processor for—in said particular adaptation-mode of operation—receiving optimized node parameters, and to apply said optimized node parameters to said nodes of said neural network to thereby implement an optimized neural network in said neural network processor, wherein the optimized node parameters have been selected among a multitude of sets of node parameters for respective candidate neural networks according to a predefined criterion in dependence of said feature vectors. A method of selecting optimized parameters for a neural network for use in a portable hearing device is further disclosed. The invention may e.g. be used in hearing aids or headsets, or similar, e.g. wearable, devices.

To provide improved feedback reduction in hearing assistance devices, technical solutions described herein include remeasuring a feedback path and updating adaptive feedback cancellation parameters whenever a user receives and plays an audio stream signal. When the user converts the audio stream signal into an acoustic audio signal using a speaker within the hearing assistance device, a feedback portion of the acoustic audio signal may be fed back into the microphone of the hearing assistance device. The hearing assistance device can then compare the feedback portion to the received audio stream signal, and that comparison can be used to update adaptive feedback cancellation parameters within the hearing assistance device. When the hearing assistance device receives an acoustic audio input at the microphone, it may amplify that input and apply the received acoustic audio input based on the updated adaptive feedback cancellation parameters to reduce or minimize feedback.

A system for detecting parasitic oscillation in a wearable audio device that includes an electro-acoustic transducer that is configured to develop sound for a user, a housing that holds the transducer, at least one of a feedforward microphone that is configured to detect sound outside of the housing and output a feedforward microphone signal and a feedback microphone that is configured to detect sound inside of the housing and output a feedback microphone signal, and an opening in the housing that emits sound pressure from the transducer. The system includes a parasitic oscillation detector that is configured to determine a fundamental frequency of at least one of the feedforward and feedback microphone signals and compare an amplitude of the determined fundamental frequency to a threshold level, to determine parasitic oscillation.

A hearing aid is configured to be worn in and/or at an ear of a user, and comprises a) an input transducer for converting an input sound to an electric input signal representing sound, h) an output transducer for converting a processed electric output signal to an output sound, c) a signal processor operationally coupled to the input and output transducers and configured to apply a forward gain to the electric input signal or a signal originating therefrom, wherein the input transducer, the signal processor and the output transducer forming part of a forward path of the hearing aid. The hearing aid further comprises d) a feedback control system for compensating for acoustic or mechanical feedback of an external feedback path from the output transducer to the input transducer, wherein the feedback control system comprises i) a feedback estimation unit for providing a feedback estimate signal of said external feedback path, ii) a combination unit located in the forward path for combining the electric input signal or a signal derived therefrom and the feedback signal detected by said estimation unit, to provide a resulting feedback corrected signal, iii) a correlation detection unit configured to determine a correlation measure between said feedback corrected signal and said output signal, said correlation detection unit further configured to provide a processed version of said correlation measure.