Hearing device or system comprising a user identification unit

Patent metadata

Field	Value
Publication number	US-12417771-B2
Application number	US-202318098848-A
Country	US
Kind code	B2
Filing date	Jan 19, 2023
Priority date	Mar 13, 2019
Publication date	Sep 16, 2025
Grant date	Sep 16, 2025

Abstract

A hearing system comprises a hearing device, e.g., a hearing aid, comprising at least one microphone for converting a sound in the environment to an electric input signal. The hearing system comprises a processor comprising a user identification unit comprising a data-driven algorithm configured to provide a user identification signal indicating whether or not, or with what probability, the person currently wearing the hearing device is a particular user in dependence of a time segment of said particular user's voice and said at least one electric input signal.

First claim

The invention claimed is:

1. A hearing system comprising
a hearing device configured to be worn by a user at or in an ear, the hearing device comprising
at least one microphone located in a behind-the-ear (BTE) part of the hearing device for converting a sound in the environment of the hearing device to at least one electric input signal,

a feedback estimation unit for providing, based on said at least one electric input signal, an estimate of a current feedback path from an output transducer of the hearing device to said at least one microphone,
a processor comprising a user identification unit, the user identification unit comprising a data-driven algorithm configured to provide a user identification signal indicating whether or not, or with what probability, the person currently wearing the hearing device is a particular user in dependence of said estimate of the current feedback path.

2. A hearing system according to claim 1 wherein said user identification unit is configured to determine whether or not or with what probability the voice of said person currently wearing the hearing device matches a voice of said particular user and to provide said user identification signal indicative thereof.

3. A hearing system according to claim 1 comprising at least two microphones located in the BTE part for providing at least two electric input signals, and at least two feedback estimation units configured to provide respective estimates of current feedback paths from the output transducer to said respective at least two microphones,
wherein the user identification unit provides the user identification signal by computing a difference between two of the estimates of current feedback paths.

4. A hearing system according to claim 3 comprising a directional microphone system for providing an own voice beamformer based on predetermined or adaptively updated own voice filter weights, wherein an estimate of the own voice of the person currently wearing the hearing device is provided in dependence of said own voice filter weights and said at least two electric input signals.

5. A hearing system according to claim 3 comprising first and second hearing devices adapted for being located at or in first and second ears, respectively, of the user, each of the first and second hearing devices comprising at least one of said at least two microphones.

6. A hearing system according to claim 1 wherein the data driven algorithm comprises a neural network.

7. A hearing system according to claim 6 configured to provide that the neural network is or has been trained based on said time segment of the particular user's voice while the particular user wears the hearing device.

8. A hearing system according to claim 7 wherein said the data-driven algorithm is or has been trained based on one-shot learning using relatively few examples of relatively short duration to learn from.

9. A hearing system according to claim 7 wherein said neural network is or has been trained based on a single time segment of the user's voice.

10. A hearing system according to claim 9 wherein said single time segment of the user's voice is of less than 20 sec. duration.

11. A hearing system according to claim 6 wherein the neural network comprises a Siamese network to learn features for each person's voice, such that a distance measure between voice features of the same person is small, while the distance between voice features of different persons is much higher.

12. A hearing system according to claim 1 configured to be brought into an authorizing mode, when said user identification signal indicates a match with the particular user.

13. A hearing system according to claim 12 configured to stay in the authorizing mode until 1) the user identification signal does not indicate a match with the particular user's identity, or 2) a request from the user is received by the hearing system, or 3) a particular termination criterion is fulfilled, or a combination thereof.

14. A hearing system according to claim 13 wherein said particular termination criterion is related to an estimate of a current feedback path of the hearing device.

15. A hearing system according claim 12 configured to enable or disable functionality of the hearing device in dependence of being in said authorizing mode.

16. A hearing system according to claim 1 wherein the hearing device is constituted by or comprises a hearing aid, a headset, an earphone, an ear protection device, an ear bud, or a combination thereof.

17. A hearing system comprising
a hearing device configured to be worn by a user at or in an ear, the hearing device comprising
at least one microphone for converting a sound in the environment of the hearing device to an electric input signal,

an own voice acoustic channel analyzer for estimating characteristics of an acoustic channel from the mouth of a user presently wearing the hearing device and the at least one microphone, based on the electric signal and a known placement of the at least one microphone relative to the user's mouth, and
a processor comprising a user identification unit, the user identification unit being configured to provide a user identification signal indicating whether or not, or with what probability, the person currently wearing the hearing device is a particular user in dependence of the estimated characteristics of the acoustic channel.

18. A hearing system according to claim 17 comprising at least two microphones for providing at least two electric input signals, and at least two own voice acoustic channel analyzers configured to estimate characteristics of respective acoustic channels from the mouth of the user to said respective at least two microphones,
wherein user identification unit is configured to provide the user identification signal based on a computed difference between two of the acoustic channels.

19. A method comprising training a data-driven algorithm for a hearing device configured to be worn by a user at or in an ear, the hearing device comprising at least one microphone located in a behind-the-ear part of the hearing device for converting a sound in the environment of the hearing device to at least one electric input signal, the data-driven algorithm being configured to provide a user identification signal indicating whether or not, or with what probability, a person currently wearing the hearing device is a particular user in dependence of an estimate of a current feedback path from an output transducer of the hearing device to said at least one microphone based on said at least one electric input signal.

20. A method according to claim 19 wherein said training of the data-driven algorithm is performed in advance of use of the hearing device by the particular user.

21. A method according to claim 19 wherein said data-driven algorithm comprises a neural network.

22. A method according to claim 21 wherein said neural network is trained based on a time segment of the particular user's voice while the particular user wears the hearing device.

23. A method according to claim 21 wherein the neural network comprises a fully connected network with two hidden layers.

24. A method according to claim 21 wherein the neural network comprises a convolution neural network comprising two layers with 1D convolution stages in combination with a max-pooling function linked to a fully connected network comprising 2 hidden layers.

25. A method accord

Assignees

• Oticon As

Inventors

• Kwasiborski Sławomir
• Guld Morten
• Petersen Svend Oscar
• Pedersen Michael Syskind
• Lesimple Christophe

Classifications

• H04R2225/41

  Detection or adaptation of hearing aid parameters or programs to listening situation, e.g. pub, forest · CPC title

• H04R2203/12

  Beamforming aspects for stereophonic sound reproduction with loudspeaker arrays · CPC title

• H04R25/43

  Electronic input selection or mixing based on input signal analysis, e.g. mixing or selection between microphone and telecoil or between microphones with different directivity characteristics (H04R25/407 takes precedence) · CPC title

• H04R25/405

  by combining a plurality of transducers · CPC title

• G10L17/18

  Artificial neural networks; Connectionist approaches · CPC title