HEARING AID APPARATUS WITH FUNDAMENTAL FREQUENCY MODIFICATION

• US 20160261959A1
• Filed: 05/16/2016
• Published: 09/08/2016
• Est. Priority Date: 11/28/2013
• Status: Active Grant

• Current Assignee: Fraunhofer Gesellschaft Zur Foerderung Der Angewandten Forsching E.V.
• Original Assignee: Fraunhofer Gesellschaft Zur Foerderung Der Angewandten Forsching E.V.
• Inventors: Harczos, Tamas
• Granted Patent: US 9,936,308 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: A61N 1/36038 Cochlear stimulation
• CPC Class Codes: G10L 21/013 Adapting to target pitch
• CPC Class Codes: H04R 2225/43 Signal processing in hearing aids to enhance the speech intelligibility
• CPC Class Codes: H04R 25/353 Frequency, e.g. frequency shift or compression

Abstract

A hearing aid apparatus includes a frequency analysis device configured to determine an instantaneous fundamental frequency value of a speech signal for a time portion of the speech signal. A statistical evaluation device is configured to determine an average fundamental frequency value of the speech signal over several time portions. A hearing aid apparatus further includes a fundamental frequency modifier that is configured to modify the instantaneous fundamental frequency value to a modified fundamental frequency value such that a difference or a quotient of the instantaneous fundamental frequency value is changed to the average fundamental frequency value according to a specific function. Thereby, a frequency range may be modified within which the fundamental frequency value varies. The hearing aid apparatus further includes a speech signal generator that is configured to generate, on the basis of the modified fundamental frequency value, a speech signal modified with regard to the fundamental frequency.

Claims

1. A hearing aid apparatus, comprising:

a frequency analysis device that is configured to determine an instantaneous fundamental frequency value of a speech signal for a time portion of the speech signal;

a statistical evaluation device that is configured to determine an average fundamental frequency value of the speech signal over several time portions;

a fundamental frequency modifier that is configured to modify the instantaneous fundamental frequency value to a modified fundamental frequency value such that a difference or a quotient of the instantaneous fundamental frequency value is changed to the average fundamental frequency value according to a specific function, in order to thus modify a frequency range within which the fundamental frequency value varies; and

a speech signal generator that is configured to generate, on the basis of the modified fundamental frequency value, a speech signal modified with regard to the fundamental frequency.

2. The hearing aid apparatus according to claim 1, further comprising a device for classifying voiced time portions and unvoiced time portions, the frequency analysis device and the statistical evaluation device being configured to determine the instantaneous fundamental frequency value and the average fundamental frequency value by means of time portions of the speech signal that are classified as voiced.

3. The hearing aid apparatus according to claim 1, wherein the speech signal generator is configured to not substantially change formant frequencies of the speech signal.

4. The hearing aid apparatus according to claim 1, wherein the fundamental frequency modifier is configured to keep an average modified fundamental frequency value substantially identical with the average fundamental frequency value.

5. The hearing aid apparatus according to claim 1, wherein the frequency analysis device is part of a device for linear predictive coding analysis (LPC) and the speech signal generator is a device for linear predictive coding synthesis.

6. The hearing aid apparatus according to claim 1, wherein the speech signal generator is based on a Fast Fourier Transformation or PSOLA (pitch synchronous overlap and add).

7. The hearing aid apparatus according to claim 1, wherein the specific function by means of which the difference or the quotient is changed is a linear function comprising a proportionality factor (PRF), so that the difference or the quotient is scaled by means of the proportionality factor.

8. The hearing aid apparatus according to claim 1, wherein the fundamental frequency modifier is configured to limit the modified fundamental frequency value at at least one of a lower limit value and an upper limit value.

9. The hearing aid apparatus according to claim 1, wherein the statistical evaluation device is configured to determine a temporal constancy of the fundamental frequency values of several time portions and to transmit the average fundamental frequency value to the fundamental frequency modifier only when the temporal constancy is above a minimum value.

10. The hearing aid apparatus according to claim 1, further comprising a speaker change detector that is configured to detect when a change from a first speaker to another speaker has occurred within the speech signal, the statistical evaluation device and the fundamental frequency modifier being configured to interrupt data processing for the first speaker until the speaker change detector detects a change back to the first speaker.

11. A method for processing a speech signal, comprising:

determining an instantaneous fundamental frequency value of a speech signal for a time portion of the speech signal;

determining an average fundamental frequency value of the speech signal over several time portions;

modifying the instantaneous fundamental frequency value to a modified fundamental frequency value such that a difference or a quotient of the instantaneous fundamental frequency value is changed to the average fundamental frequency value according to a specific function in order to modify a frequency range within which the fundamental frequency value varies; and

generating, on the basis of the modified fundamental frequency value, a speech signal modified with regard to the fundamental frequency.

12. A non-transitory digital storage medium having a computer program stored thereon to perform a method for processing a speech signal, said method comprising:

determining an instantaneous fundamental frequency value of a speech signal for a time portion of the speech signal;

determining an average fundamental frequency value of the speech signal over several time portions;

modifying the instantaneous fundamental frequency value to a modified fundamental frequency value such that a difference or a quotient of the instantaneous fundamental frequency value is changed to the average fundamental frequency value according to a specific function in order to modify a frequency range within which the fundamental frequency value varies; and

generating, on the basis of the modified fundamental frequency value, a speech signal modified with regard to the fundamental frequency,

when said computer program is run by a computer.