US 12,283,286 B2
Signal processing device, system and method for processing audio signals
Peter Udo Diehl, Berlin (DE); and Elias Sprengel, Berlin (DE)
Assigned to Sonova AG, Staefa (CH)
Appl. No. 17/425,256
Filed by Sonova AG, Staefa (CH)
PCT Filed Jan. 24, 2020, PCT No. PCT/EP2020/051735
§ 371(c)(1), (2) Date Jul. 22, 2021,
PCT Pub. No. WO2020/152324, PCT Pub. Date Jul. 30, 2020.
Claims priority of application No. 10 2019 200 956.5 (DE), filed on Jan. 25, 2019.
Prior Publication US 2022/0093118 A1, Mar. 24, 2022
Int. Cl. G10L 25/30 (2013.01); G06N 3/044 (2023.01); G06N 3/045 (2023.01); G10L 21/0272 (2013.01); H04R 25/00 (2006.01)
CPC G10L 25/30 (2013.01) [G06N 3/044 (2023.01); G06N 3/045 (2023.01); G10L 21/0272 (2013.01); H04R 25/507 (2013.01)] 19 Claims
OG exemplary drawing
 
1. A signal processing device for processing audio signals, having:
an input interface for receiving an input signal;
a neural conditioning network that is configured to condition the input signal;
a plurality of neural networks for separating one or more audio signals from the input signal output from the neural conditioning network, wherein each neural network included in the plurality of neural networks is customized to a different type of audio signal; and
an output interface for outputting an output signal, wherein the output signal comprises the one or more audio signals separated from the input signal,
wherein:
the number of audio signals separated from the input signal by using at least one neural network included in the plurality of neural networks is variable; and
the number of neural networks included in the plurality of neural networks used for separating the one or more audio signals from the input signal is variable depending on the input signal and at least one of a battery charge state or available computing power.
 
6. A system for processing audio signals, having:
at least one signal processing device for processing audio signals according to claim 1;
at least one recording device for recording an input signal, wherein the at least one recording device is connected to the at least one signal processing device via the input interface in a data-transmitting manner; and
at least one playback device for playing back an output signal, wherein the at least one playback device is connected to the at least one signal processing device via the output interface in a data-transmitting manner.
 
12. A method for processing audio signals, having the steps of:
providing a signal processing device according to claim 1;
providing an input signal;
supplying the input signal to the signal processing device via the input interface;
separating one or more audio signals from the input signal by using the plurality of neural networks, wherein the number of audio signals separated from the input signal by using the plurality of neural networks is variable;
determining a priority parameter for each of the one or more audio signals;
modulating each audio signal included in the one or more audio signals based on the respective priority parameter;
combining the audio signals to produce an output signal; and
outputting the output signal via the output interface.