Criteria for usable repetitions of an utterance during speech reference enrollment

• US 6,012,027 A
• Filed: 09/17/1997
• Issued: 01/04/2000
• Est. Priority Date: 05/27/1997
• Status: Expired due to Term

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: Ameritech Corp. (AT&T, Inc.)
• Inventors: Bossemeyer, Jr., Robert Wesley
• Primary Examiner(s): Hudspeth, David R.
• Assistant Examiner(s): Storm, Donald L.
• Application Number:
  US08/932,078
• Time in Patent Office: 839 Days
• Field of Search: 704/243, 704/238, 704/239, 704/248, 704/253, 704/251, 704/231
• US Class Current: 704/243
• CPC Class Codes: G10L 15/07 to the speaker
• CPC Class Codes: G10L 17/04 Training, enrolment or model building
• CPC Class Codes: G10L 2015/0631 Creating reference templates; Clustering
• CPC Class Codes: G10L 2015/0636 Threshold criteria for the updating
• CPC Class Codes: G10L 2015/0638 Interactive procedures
• CPC Class Codes: H04M 2201/40 using speech recognition speech recognition per se G10L15/00
• CPC Class Codes: H04M 3/382 using authorisation codes or passwords
• CPC Class Codes: H04M 3/493 Interactive information services, e.g. directory enquiries ; Arrangements therefor, e.g. interactive voice response [IVR] systems or voice portals

Abstract

A speech reference enrollment method involves the following steps: (a) requesting a user speak a vocabulary word; (b) detecting a first utterance (354); (c) requesting the user speak the vocabulary word; (d) detecting a second utterance (358); (e) determining a first similarity between the first utterance and the second utterance (362); (f) when the first similarity is less than a predetermined similarity, requesting the user speak the vocabulary word; (g) detecting a third utterance (366); (h) determining a second similarity between the first utterance and the third utterance (370); and (i) when the second similarity is greater than or equal to the predetermined similarity, creating a reference (364).

Claims

1. A speech reference enrollment method, comprising the steps of:

(a) receiving a first utterance of a vocabulary word;

(b) extracting a plurality of features from the first utterance;

(c) receiving a second utterance of the vocabulary word;

(d) determining a duration of the second utterance;

(e) when the duration is less than a minimum duration, requesting a user speak a third utterance of the vocabulary word and proceeding to step (i);

(f) extracting the plurality of features from the second utterance;

(g) determining a first similarity between the plurality of features from the first utterance and the plurality of features from the second utterance;

(h) when the first similarity is less than a predetermined similarity, requesting a user to speak a third utterance of the vocabulary word;

(i) extracting the plurality of features from the third utterance;

(j) determining a second similarity between the plurality of features from the first utterance and the plurality of features from the third utterance; and

(k) when the second similarity is greater than or equal to the predetermined similarity, forming a reference for the vocabulary word.

2. The method of claim 1, further including the steps of:

(l) when the second similarity is less than the predetermined similarity, determining a third similarity between the plurality of features from the second utterance and the plurality of features from the third utterance;

(m) when the third similarity is greater than or equal to the predetermined similarity, forming the reference for the vocabulary word.

3. The method of claim 2, further including the steps of:

(n) when the third similarity is less than the predetermined similarity, returning to step (a).

4. The method of claim 1, wherein step (c) further includes the steps of:

(c1) determining a duration of the second utterance;

(c2) when the duration is greater than a maximum duration, disregarding the second utterance.

5. The method of claim 4, wherein step (c1) further includes the steps of:

(i) setting an amplitude threshold;

(ii) determining a start time when an input signal exceeds the amplitude threshold;

(iii) determining an end time, after the start time, when the input signal is less than the amplitude threshold;

(iv) calculating the duration as a difference between the end time and the start time.

6. The method of claim 1, wherein step (f) further includes the steps of:

(f1) determining an estimate of a number of voiced speech frames;

(f2) when the estimate of the number of voiced speech frames is less than a threshold requesting the user repeat the vocabulary word;

(f3) returning to step (c).

7. The method of claim 1, wherein step (a) further includes the steps of:

(a1) determining a signal to noise ratio of the first utterance;

(a2) when the signal to noise ratio is less than a predetermined signal to noise ratio, increasing a gain of a voice amplifier.

8. The method of claim 7, further including the step of:

(a3) requesting the user repeat the vocabulary word.

9. The method of claim 1, wherein step (b) further includes the step of:

(b1) determining an amplitude histogram of the first utterance.

10. A speech reference enrollment method, comprising the steps of:

(a) requesting a user speak a vocabulary word;

(b) detecting a first utterance;

(c) determining if the first utterance exceeds an amplitude threshold;

(d) when the first utterance does not exceed the amplitude threshold, return to step (a);

(e) requesting the user speak the vocabulary word;

(f) detecting a second utterance;

(g) determining a first similarity between the first utterance and the second utterance;

(h) when the first similarity is less than a predetermined similarity, requesting the user speak the vocabulary word;

(i) detecting a third utterance;

(i) determining a second similarity between the first utterance and the third utterance; and

(k) when the second similarity is greater than or equal to the predetermined similarity, creating a reference.

11. The method of claim 10, further including the steps of:

(l) determining a third similarity between the second utterance and the third utterance;

(m) when the third similarity is greater than or equal to the predetermined similarity, creating the reference.

12. The method of claim 11, further including the steps of:

(n) when the third similarity is less than the predetermined similarity, returning to step (a).

13. The method of claim 10, wherein step (b) further includes the steps of:

(b1) determining an estimate of a number of voiced speech frames;

(b2) when the number of voiced speech frames is less than a predetermined number of voiced speech frames, returning to step (a).

14. The method of claim 10, wherein step (b) further includes the steps of:

(b1) determining a duration of the first utterance;

(b2) when the duration is less than a minimum duration, returning to step (a);

(b3) when the duration is greater than a maximum duration, returning to step (a).

15. A computer readable storage medium containing computer readable instructions that when executed by a computer performs the following steps:

(a) requesting a user speak a vocabulary word;

(b) receiving a first digitized utterance;

(c) extracting a plurality of features from the first digitized utterance;

(d) determining a signal to noise ratio;

(e) when the signal to noise ratio is less than a predetermined signal to noise ratio, returning to step (a);

(f) requesting the user speak the vocabulary word;

(g) receiving a second digitized utterance of the vocabulary word;

(h) extracting the plurality of features from the second digitized utterance;

(i) determining a first similarity between the plurality of features from the first digitized utterance and the plurality of features from the second digitized utterance;

(j) when the first similarity is less than a predetermined similarity, requesting the user to speak a third utterance of the vocabulary word;

(k) extracting the plurality of features from a third digitized utterance;

(l) determining a second similarity between the plurality of features from the first digitized utterance and the plurality of features from the third digitized utterance; and

(m) when the second similarity is greater than or equal to the predetermined similarity, forming a reference for the vocabulary word.

16. The computer readable storage medium of claim 15, further executing the steps of:

(n) when the second similarity is less than the predetermined similarity, determining a third similarity between the plurality of features from the second digitized utterance and the plurality of features from the third digitized utterance;

(o) when the third similarity is greater than or equal to the predetermined similarity, forming the reference for the vocabulary word.

17. The computer readable storage medium of claim 16, further executing the steps of:

(p) when the third similarity is less than the predetermined similarity, returning to step (a).

18. The computer readable storage medium of claim 15, wherein step (e) further includes the steps of:

(e1) determining if an amplifier gain is saturated;

(e2) when the amplifier gain is saturated, going to step (a).

19. The computer readable storage medium of claim 15, wherein step (e) further includes the step of increasing a gain of an amplifier.

20. The computer readable storage medium of claim 18, wherein step (e2) further includes the step of decreasing a gain of an amplifier.