Systems and methods for estimating age of a speaker based on speech

Patent metadata

Field	Value
Publication number	US-10269356-B2
Application number	US-201615243635-A
Country	US
Kind code	B2
Filing date	Aug 22, 2016
Priority date	Aug 22, 2016
Publication date	Apr 23, 2019
Grant date	Apr 23, 2019

Abstract

There is provided a system comprising a microphone, configured to receive an input speech from an individual, an analog-to-digital (A/D) converter to convert the input speech to digital form and generate a digitized speech, a memory storing an executable code and an age estimation database, a hardware processor executing the executable code to receive the digitized speech, identify a plurality of boundaries in the digitized speech delineating a plurality of phonemes in the digitized speech, extract a plurality of formant-based feature vectors from each phoneme in the digitized speech based on at least one of a formant position, a formant bandwidth, and a formant dispersion, compare the plurality of formant-based feature vectors with age determinant formant-based feature vectors of the age estimation database, determine the age of the individual when the comparison finds a match in the age estimation database, and communicate an age-appropriate response to the individual.

First claim

What is claimed is:

1. A system comprising:
a microphone configured to receive an input speech from an individual;
an analog-to-digital (A/D) converter configured to convert the input speech from an analog form to a digital form and generate a digitized speech;
a memory storing an executable code and an age estimation database including a plurality of age determinant formant-based feature vectors;
a hardware processor executing the executable code to:
receive the digitized speech from the A/D converter;
identify a plurality of boundaries between a plurality of phonemes in the digitized speech;
extract a plurality of formant-based feature vectors from one or more phonemes of the plurality of phonemes delineated by the plurality of boundaries, based on a formant position, a formant bandwidth, and a formant dispersion, wherein the formant dispersion is a geometric mean of the formant spacing;
compare the plurality of formant-based feature vectors with the age determinant formant-based feature vectors of the age estimation database;
estimate the age of the individual when the comparison finds a match in the age estimation database; and
communicate an age-appropriate response to the individual based on the estimated age of the individual.

2. The system of claim 1 , wherein estimating the age of the individual includes a weighted combination of two or more age determinant formant-based feature vectors of the plurality of age determinant formant-based feature vectors.

3. The system of claim 1 , wherein the input speech is one of a predetermined sequence of phonemes and natural speech.

4. The system of claim 1 , wherein the age determinant formant-based feature vectors of the age estimation database include a plurality of formant-based feature vectors corresponding to a plurality of most predictive phonemes, wherein each of the plurality of most predictive phonemes corresponds to a different age.

5. The system of claim 1 , wherein the digitized speech includes at least one of a silence and a filled pause.

6. The system of claim 1 , wherein the input speech includes a plurality of formants where each formant of the plurality of formants is a resonance of a vocal tract of the individual.

7. The system of claim 1 , wherein the input speech is one of English and a language that is not English.

8. The system of claim 1 , wherein the age of the individual is estimated probabilistically.

9. A method for use with a system having a microphone, an analog-to-digital (A/D) converter, a memory storing an executable code, and a hardware processor, the method comprising:
receiving, using the hardware processor, a digitized speech from the A/D converter;
identifying, using the hardware processor, a plurality of boundaries between a plurality of phonemes in the digitized speech;
extracting, using the hardware processor, a plurality of formant-based feature vectors from one or more phonemes of the plurality of phonemes delineated by the plurality of boundaries, based on a formant position, a formant bandwidth, and a formant dispersion, wherein the formant dispersion is a geometric mean of the formant spacing;
comparing, using the hardware processor, the plurality of formant-based feature vectors with the age determinant formant-based feature vectors of the age estimation database;
estimating, using the hardware processor, the age of the individual when the comparison finds a match in the age estimation database; and
communicating, using the hardware processor, an age-appropriate response to the individual based on the estimated age of the individual.

10. The method of claim 9 , wherein estimating the age of the individual includes a weighted combination of two or more age determinant formant-based feature vectors of the plurality of age determinant formant-based feature vectors.

11. The method of claim 9 , wherein the input speech is one of a predetermined sequence of phonemes and natural speech.

12. The method of claim 9 , wherein the age determinant formant-based feature vectors of the age estimation database include a plurality of formant-based feature vectors corresponding to a plurality of most predictive phonemes, wherein each of the plurality of most predictive phonemes corresponds to a different age.

13. The method of claim 9 , wherein the digitized speech includes at least one of a silence and a filled pause.

14. The method of claim 9 , wherein the input speech includes a plurality of formants where each formant of the plurality of formants is a resonance of a vocal tract of the individual.

15. The method of claim 9 , wherein the input speech is one of English and a language that is not English.

16. The method of claim 9 , wherein the age of the individual is estimated probabilistically.

17. The system of claim 1 , wherein prior to the extracting of the plurality of formant-based feature vectors, the hardware processor executes the executable code to identify a segment of one or more phonemes of the plurality of phonemes delineated by the plurality of boundaries, and wherein the extracting extracts the plurality of formant-based feature vectors from the identified segment of the one or more phonemes of the plurality of phonemes delineated by the plurality of boundaries.

18. The method of claim 9 , wherein prior to the extracting of the plurality of formant-based feature vectors, the method further comprises:
identifying a segment of one or more phonemes of the plurality of phonemes delineated by the plurality of boundaries;
wherein the extracting extracts the plurality of formant-based feature vectors from the identified segment of the one or more phonemes of the plurality of phonemes delineated by the plurality of boundaries.

Assignees

• Disney Entpr Inc

Inventors

• Singh Rita
• Lehman Jill Fain

Classifications

• G10L17/02

  Preprocessing operations, e.g. segment selection; Pattern representation or modelling, e.g. based on linear discriminant analysis [LDA] or principal components; Feature selection or extraction · CPC title

• G10L15/10

  using distance or distortion measures between unknown speech and reference templates · CPC title

• G10L17/26 Primary

  Recognition of special voice characteristics, e.g. for use in lie detectors; Recognition of animal voices · CPC title

• G10L15/08

  Speech classification or search · CPC title

• G10L25/51

  for comparison or discrimination · CPC title