Speaker and command verification method

• US 5,835,894 A
• Filed: 02/14/1997
• Issued: 11/10/1998
• Est. Priority Date: 01/19/1995
• Status: Expired due to Term

• Current Assignee: Voice Security Systems Incorporated
• Original Assignee: Ann Adcock Corporation
• Inventors: Schwartz, Gerald S., Adcock, Sherrie Ann
• Primary Examiner(s): Dorvil, Richemond
• Assistant Examiner(s): Chawan, Vijay B.
• Application Number:
  US08/800,354
• Time in Patent Office: 634 Days
• Field of Search: 704/200, 704/251, 704/253, 704/254, 704/246, 704/242, 704/273, 704/500, 704/501, 704/502, 704/503, 704/275, 704/508, 704/274
• US Class Current: 704/273
• CPC Class Codes: G10L 17/00 Speaker identification or verification techniques
• CPC Class Codes: G10L 17/04 Training, enrolment or model building
• CPC Class Codes: G10L 17/22 Interactive procedures; Man-machine interfaces
• CPC Class Codes: G10L 17/24 the user being prompted to utter a password or a predefined phrase

Abstract

A security method compares a present verbal utterance with a previously recorded verbal utterance by comparing frequency domain representations of the utterances, with multiple repeat utterances forming a basis for determining a variation in repetitious performance by an individual, and similar differences between enrollment and challenge utterances forming a basis for a similar analysis of variance between enrollment and challenge utterances. In one embodiment a set of enrollment data is searched by each challenge until either a match is made, indicating an action, possibly dependent upon the specific match, or no match is made indicating an abort.

Claims

1. A method of comparing a challenge verbal utterance with an enrollment verbal utterance to determine a security action, the method comprising the steps of:

A) preparing an enrollment data comprising the steps of;

a) converting an enrollment verbal utterance into a first electrical signal;

b) creating a fast Fourier transformation of the signal to produce a frequency domain representation of the signal;

c) taking M samples of the signal for each of N frequency channels to form an M by N sample enrollment matrix Ei;

d) determining if at least X of the M samples have a selected minimum energy content in at least Y of the N frequency channels;

e) repeating steps (a) through (d) if step (d) fails, and aborting after a selected number of repeats of steps (a) through (d);

f) repeating steps (a) through (d) if step (d) succeeds, until a selected number of said sample matrices Ei have been formed as E1, E2, E3, . . . . ;

g) forming the difference between each pair of the enrollment matrices Ei, as (E1-E2), (E1-E3), (E2-E3), . . . ;

h) forming the sum, S, of the differences between each of the pairs of the matrices Ei, as S={(E1-E2)+(E1-E3)+(E1-E2)+. . . };

i) accepting S as an enrollment data if S<

EB, EB being a selected first enrollment criterion and also if each said difference between each pair of the matrices Ei is greater than ES, ES being a selected second enrollment criterion;

B) challenging the enrollment data comprising the steps of;

a) converting a challenge verbal utterance into a second electrical signal;

b) creating a second fast Fourier transformation of the second signal to produce a frequency domain representation of the second signal;

c) taking M samples of the second signal for each of N frequency channels to form an M by N sample challenge matrix C;

d) determining if at least X of the M samples have a selected minimum energy content in at least Y of the N frequency channels;

e) repeating steps (a) through (d) if step (d) fails, and aborting after a selected number of repeats of steps (a) through (d);

f) forming the difference between C and Di for each of the matrices Di, as (C-D1), (C-D2), (C-D3), . . . ;

g) forming the sum, S'"'"', of the differences between each of the pairs C and Di, as S'"'"'={(C-D1)+(C-D2)+(C-D3)+. . . };

h) accepting S'"'"' as a valid challenge if S'"'"'<

VB, VB being a selected first verification criterion and also if each said difference between each of the pairs C and Di is greater than VS, VS being a selected second verification criterion;

C) enabling the security action if the challenge verbal utterance is accepted.

2. A method of comparing a challenge verbal utterance with a set of enrollment verbal utterances to determine a security action, the method comprising the steps of:

A) preparing enrollment data comprising the steps of;

a) converting an enrollment verbal utterance into a first electrical signal;

b) creating a fast Fourier transformation of the signal to produce a frequency domain representation of the signal;

c) taking M samples of the signal for each of N frequency channels to form an M by N sample enrollment matrix Ei;

d) determining if at least X of the M samples have a selected minimum energy content in at least Y of the N frequency channels;

e) repeating steps (a) through (d) if step (d) fails, and aborting after a selected number of repeats of steps (a) through (d);

f) repeating steps (a) through (d) if step (d) succeeds, until a selected number of said sample matrices Ei have been formed as E1, E2, E3 . . . ,

g) forming the difference between each pair of the enrollment matrices Ei, as (E1-E2), (E1-E3), (E2-E3), . . . ;

h) forming the sum, S, of the differences between each of the pairs of the matrices Ei, as S={(E1-E2)+(E1-E3)+(E1-E2)+. . . };

i) accepting S as an enrollment data if S<

EB, EB being a selected first enrollment criterion and also if each said difference between each pair of the matrices Ei is greater than ES, ES being a selected second enrollment criterion;

j) repeating steps (a) to (i) for a plurality of individual enrollment data sets;

B) challenging the enrollment data, the enrollment challenging comprising the steps of;

a) converting a challenge verbal utterance into a second electrical signal;

b) creating a second fast Fourier transformation of the second signal to produce a frequency domain representation of the second signal;

c) taking M samples of the second signal for each of N frequency channels to form an M by N sample challenge matrix C;

d) determining if at least X of the M samples have a selected minimum energy content in at least Y of the N frequency channels;

e) repeating steps (a) through (d) if step (d) fails, and aborting after a selected number of repeats of steps (a) through (d);

f) forming the difference between C and Di for each of the matrices Di, as (C-D1), (C-D2), (C-D3), . . . ;

g) forming the sum, S'"'"', of the differences between each of the pairs C and Di, as S'"'"'={(C-D1)+(C-D2)+(C-D3)+. . . };

h) repeating steps (a) through (g) for each of the enrollment data sets until S'"'"'<

VB, VB being a selected first verification criterion, and each said difference between each of the pairs C and Di is greater than VS, VS being a selected second verification criterion;

i) aborting the method when step (h) has been completed without the challenge verbal utterance being accepted;

C) enabling the security action if the challenge verbal utterance is accepted in step (B)(h).

3. The method of claim 2 wherein each of the data sets includes data defining a specific security action wherein upon completion of step (B)(h) a further step is taken to complete said specific security action, said action including the generating of an electrical signal.