Ear-worn device with neural network for noise reduction and/or spatial focusing using multiple input audio signals
An ear-worn device may include two or more microphones configured to generate time-domain audio signals, each of the two or more microphones configured to generate one of the time-domain audio signals; processing circuitry comprising analog processing circuitry, digital processing circuitry, beamforming circuitry, and short-time Fourier transformation (STFT) circuitry, the processing circuitry configured to generate, from the time-domain audio signals, one or more frequency-domain non-beamformed audio signals and one or more frequency-domain beamformed signals; and enhancement circuitry comprising neural network circuitry configured to receive multiple frequency-domain input audio signals originating from the one or more frequency-domain non-beamformed audio signals and the one or more frequency-domain beamformed signals, and implement a single neural network trained to generate, based on the multiple frequency-domain input audio signals, a noise-reduced and spatially-focused output audio signal or an output for generating a noise-reduced and spatially-focused output audio signal.
1. An ear-worn device, comprising:
two or more microphones configured to generate audio signals, each of the two or more microphones configured to generate one of the audio signals;
wherein the two or more microphones include a front microphone and a back microphone, and
wherein the front microphone is configured to generate a front time-domain audio signal and the back microphone is configured to generate a back time-domain audio signal;
processing circuitry comprising beamforming circuitry, the processing circuitry configured to generate, from the audio signals, one or more non-beamformed audio signals originating from the front time-domain audio signal and the back time-domain audio signal and one or more beamformed audio signals, the one or more non-beamformed audio signals each comprising a speech portion and a noise portion, and the one or more beamformed audio signals each comprising a speech portion and a noise portion; and
neural network circuitry downstream of the beamforming circuitry, the neural network circuitry configured to:
receive multiple input signals originating from the one or more non-beamformed audio signals and the one or more beamformed audio signals; and
implement one or more neural networks trained to generate, based on the multiple input signals:
a noise-reduced and/or spatially-focused output audio signal; or
an output configured to generate the noise-reduced and/or spatially-focused output audio signal.
2. The ear-worn device of claim 1 , wherein the two or more microphones comprise exactly two or exactly three microphones.
3. The ear-worn device of claim 1 , wherein:
the output audio signal is spatially focused; and
the generating of the output audio signal or the output configured to generate the output audio signal is based on a wearer selection of a size of a front-facing spatial region.
4. A system comprising:
the ear-worn device of claim 3 ; and
a processing device in communication with the ear-worn device and configured to receive the wearer selection of the size of the front-facing spatial region.
5. The system of claim 4 , wherein the processing device is configured to display multiple options for the size of the front-facing spatial region.
6. The system of claim 5 , wherein the processing device is configured to display exactly two, exactly three, or exactly four options for the size of the front-facing spatial region.
7. The system of claim 5 , wherein the processing device is configured, when displaying the multiple options for the size of the front-facing spatial region, to display graphical representations of the multiple options for the size of the front-facing spatial region.
8. The ear-worn device of claim 3 , wherein the output audio signal uses a mapping of gains to respective spatial regions relative to a wearer of the ear-worn device.
9. The ear-worn device of claim 8 , wherein the mapping of the gains to the respective spatial regions comprises mapping more than two spatial regions each to a different gain, and one or more of the spatial regions are processed with gains not equal to 1 or 0.
10. The ear-worn device of claim 1 , wherein the one or more beamformed audio signals comprise multiple beamformed audio signals each having a different directional pattern, and at least one of the multiple beamformed audio signals has a dipole, hypercardioid, supercardioid, or cardioid directional pattern.
11. The ear-worn device of claim 1 , wherein the neural network circuitry is configured to output a single output based on the multiple input signals.
12. An ear-worn device, comprising:
two or more microphones, wherein:
the two or more microphones comprise at least a front microphone and a back microphone; and
the front microphone is configured to generate a front time-domain audio signal and the back microphone is configured to generate a back time-domain audio signal;
processing circuitry comprising beamforming circuitry, the processing circuitry configured to generate, from the front time-domain audio signal and the back time-domain audio signal, multiple beamformed signals each having a different directional pattern; and
neural network circuitry downstream of the beamforming circuitry, the neural network circuitry configured to:
receive the multiple beamformed signals; and
implement one or more neural networks trained to generate, based on processing together the multiple beamformed signals, a mask configured to generate a spatially-focused output audio signal;
wherein the spatially-focused output audio signal has a mapping of gain to direction-of-arrival different from any directional pattern of the multiple beamformed signals.
13. The ear-worn device of claim 12 , wherein the two or more microphones comprise exactly two or exactly three microphones.
14. The ear-worn device of claim 12 , wherein:
the generating of the spatially-focused output audio signal is based on a wearer selection of a size of a front-facing spatial region.
15. A system comprising:
the ear-worn device of claim 14 ; and
a processing device in communication with the ear-worn device and configured to receive the wearer selection of the size of the front-facing spatial region.
16. The system of claim 15 , wherein the processing device is configured to display multiple options for the size of the front-facing spatial region.
17. The system of claim 16 , wherein the processing device is configured to display exactly two, exactly three, or exactly four options for the size of the front-facing spatial region.
18. The ear-worn device of claim 12 , wherein the mapping of gain to direction-of-arrival is relative to a wearer of the ear-worn device.
19. The ear-worn device of claim 18 , wherein the mapping of gain to direction-of-arrival comprises mapping more than two spatial regions each to a different gain, and one or more of the spatial regions are processed with gains not equal to 1 or 0.
20. The ear-worn device of claim 12 , wherein at least one of the multiple beamformed signals has a dipole, hypercardioid, supercardioid, or cardioid directional pattern.
21. The ear-worn device of claim 12 , wherein the neural network circuitry is configured to output a single output based on processing together the multiple beamformed signals.
22. The ear-worn device of claim 1 , wherein the noise-reduced and/or spatially-focused output audio signal comprises a spatially-focused output audio signal.
23. The ear-worn device of claim 22 , wherein the spatially-focused output audio signal has a mapping of gain to direction-of-arrival different from any directional pattern of the one or more beamformed audio signals.
24. The ear-worn device of claim 22 , wherein the spatially-focused output audio signal comprises speech components with different gains based on their different directions-of-arrival.
25. The ear-worn device of claim 1 , wherein the one or more beamformed audio signals comprise multiple beamformed audio signals each having a different directional pattern.
26. The ear-worn device of claim 1 , wherein the one or more beamformed audio signals comprise a front-facing beamformed signal and a back-facing beamformed signal.
27. The ear-worn device of claim 12 , wherein the spatially-focused output audio signal comprises speech components with different gains based on their different directions-of-arrival.
28. The ear-worn device of claim 12 , wherein the multiple beamformed signals comprise a front-facing beamformed signal and a back-facing beamformed signal.