Demant

A hearing aid comprising one or more sub-band filters is disclosed. The disclosure relates to the use of one or more sub-band filters in the frequency domain to attenuate background noise of a microphone signal. The disclosure relates to the use of two analysis filter banks. One of the two analysis filter banks has a low frequency resolution that outputs a set of low resolution sub-band input signals, and the other analysis filter bank has a higher frequency resolution that outputs a set of high resolution sub-band input signals. The low frequency resolution analysis filter bank has low signal delay but low frequency resolution, whereas the high frequency resolution analysis filter bank has high signal delay but high frequency resolution. The disclosure is to use the output of the high frequency resolution analysis filter bank to estimate one or more sub-band filters that are applied on the set of low resolution sub-band input signals to attenuate background noise in the set of low-resolution sub-band signals.

A method for training a neural network is disclosed. The method comprises receiving, by a sound database, a training input. The training input comprises a speech component and a noise component. The method comprises receiving, by the sound database, a training target comprising the speech component. The method comprises providing, by a high-resolution analysis filter bank, a high-resolution training sub-band input based on the training input. The method comprises determining, by the neural network comprising a plurality of neural network parameters, one or more sub-band filters based on the high-resolution training sub-band input. The method comprises providing, by a low-resolution analysis filter bank, a low-resolution training sub-band input based on the training input. The method comprises providing, by filtering the low-resolution training sub-band input using the one or more sub-band filters, a training sub-band output. The method comprises providing, by the low-resolution synthesis filter bank, a training output. The method comprises training the neural network by updating, by a network optimizer, the neural network parameters based on reducing a neural network loss between the training output and the training target.

A hearing aid comprising a dipole antenna for transmitting or receiving a wireless signal, wherein the antenna is configured to have a current amplitude maximum on a transverse antenna segment.

A hearing aid is disclosed. The hearing aid comprises a wireless interface configured to receive a wireless sound signal, and a processor configured to, initialize a silence counter in response to receiving the wireless sound signal, periodically determine a sound level based on the wireless sound signal, in accordance with the sound level being below a level threshold, increase the silence counter, and in accordance with silence counter being equal to or above a silence threshold, reset the silence counter and provide a notification to a user of the hearing aid.

A hearing aid for being worn by a user. The hearing aid comprises an input unit and a processor. The processor is configured to: obtain a hearing aid audio signal based on the input audio signal, determine a first audio parameter based on the input audio signal, obtain a first sensor signal indicative of a first biological parameter of the user, determine the first biological parameter based on the first sensor signal, determine a first correlation between the first audio parameter and the first biological parameter, and set a processing parameter of the hearing aid based on the first correlation.

A method, performed by a headphone device, is disclosed. The method comprises:- determining at least one construction characteristic of the headphone device; and- adjusting at least one sound characteristic of the headphone device based on the at least one construction characteristic determined.

A hearing aid comprising an input microphone, an output, an error microphone configured to provide an error audio signal, and a processor. The processor is configured to estimate a feedback path transfer function indicative of an audio path from the output transducer to the input microphone based on an input audio signal, an error audio signal, and an output audio signal.

A method for determining personalized training data. The comprises obtaining a set of in-situ input signals, determining a set of in-situ voice signals and a set of in-situ-noise signals, obtaining a set of generic noise signals, obtaining a set of generic clean speech signals, and determining a personalized training data set.

The invention relates to methods, devices and systems for charging of hearing devices, specifically for charging hearing aids. The hearing device is configured to be inductively charged via a loop structure and comprise: a housing, at least one microphone, a processor, a rechargeable battery, and a receiver. The hearing device advantageously includes a loop structure arranged at an outer surface part of the housing. The loop structure is configured to receive a wireless electronic signal from a charging device to charge the hearing device's comprising rechargeable battery.

A hearing aid comprising a beamformer equivalent with single-channel filtering (FB2C, BF1, GU, DIS) is presented. The disclosure may be used for noise reduction. The hearing aid comprises a) an input unit (IU) providing at a primary electric input signal (x1 (k,n)), b) a frequency band to channel unit (FB2C) providing a first secondary electric signal (x̃1 (k̃,n)) based on the primary electric input signal (x1 (k, n)), c) a wireless receiver (WR) allowing to establish a wireless link with at least another device and receiving at least a second secondary electric signal (x̃2 (k̃, n)), d) a first signal provider (SP1) for providing an output of the first signal provider based on the first secondary electric signal (x̃1 (k̃,n)) and the second secondary electric signal (x̃2 (k̃,n)), e) a second signal provider (SP2) for providing an output of the second signal provider, f) a distribution unit (DIS) distributing the input of the distribution unit, g) a processor providing a noise reduced electric signal (y1 (k,n)), and h) an output unit (OU) for providing a stimulus for being presented as an acoustic signal to the user of the hearing aid.