Method for measuring speech characteristics

• US 8,768,697 B2
• Filed: 01/29/2010
• Issued: 07/01/2014
• Est. Priority Date: 01/29/2010
• Status: Active Grant

• Current Assignee: Rosetta Stone Limited (Rosetta Stone Incorporated)
• Original Assignee: Rosetta Stone Limited (Rosetta Stone Incorporated)
• Inventors: Tepperman, Joseph, Stanley, Theban, Hacioglu, Kadri
• Primary Examiner(s): Godbold, Douglas
• Assistant Examiner(s): ESTES, ERNEST LANGSTON
• Application Number:
  US12/696,222
• Publication Number:
  US 20110191104A1
• Time in Patent Office: 1,614 Days
• Field of Search: 704/9, 704/211, 704/270, 704/236, 704/243, 434156-186
• US Class Current: 704/236
• CPC Class Codes: G09B 19/04 Speaking with audible presentation of the material to be studied G09B5/04
• CPC Class Codes: G09B 19/06 Foreign languages with audible presentation of material to be studied G09B5/04
• CPC Class Codes: G10L 15/00 Speech recognition G10L17/00 takes precedence

Abstract

In some embodiments, a method includes measuring a disparity between two speech samples by segmenting both a reference speech sample and a student speech sample into speech units. A duration disparity can be determined for units that are not adjacent to each other in the reference speech sample. A duration disparity can also be determined for the corresponding units in the student speech sample. A difference can then be calculated between the student speech sample duration disparity and the reference speech sample duration disparity.

Claims

1. A non-transitory processor-readable medium storing code representing instructions to be executed by a processor, the code comprising code to cause the processor to:

record a reference speech sample spoken by a native speaker;

record a student speech sample spoken by a language student;

generate a reference digital data file representative of the reference speech sample and a student digital data file representative of the student speech sample, the reference digital data file including information indicative of a sound intensity level as function of time over a chronological duration of the reference speech sample and the student digital data file including information indicative of a sound intensity level as a function of time over a chronological duration of the student speech sample;

segment the reference digital data file into a reference speech-unit data file and the student digital data file into a student speech-unit data file, the reference speech-unit data file including a plurality of reference speech-units from the reference digital data file over a reference speech sample duration and the student speech-unit data file including a plurality of student speech-units from the student digital data file over a student speech sample duration; and

generate a plurality of student speech-unit duration differences by calculating a difference in duration between each student speech-unit from the plurality of student speech-units and the remaining speech-units from the plurality of student speech-units.

2. The non-transitory processor-readable medium of claim 1, further comprising code to cause the processor to:

scale at least one of the reference speech sample or the student speech sample so that the total duration of the reference speech sample and the student speech sample are substantially the same.

3. The non-transitory processor-readable medium of claim 1, wherein a type of speech-unit is selected from at least one of phonemes, syllables, or words.

4. The non-transitory processor-readable medium of claim 1, further comprising code to cause the processor to:

determine the most granular level of speech in which a number of speech-units is the same for the reference speech sample and the student speech sample; and

compare a characteristic of one of the plurality of reference speech-units with the characteristic of a corresponding one of the plurality of student speech-units.

5. The non-transitory processor-readable medium of claim 1, wherein the calculating a difference in duration between each speech-unit from the plurality of student speech-units includes calculating a difference in duration between each phoneme in the student speech sample.

6. A method for measuring a disparity between two speech samples, comprising:

recording a reference speech sample spoken by a native speaker;

recording a student speech sample spoken by a language student;

in a computer process, generating a reference speech digital data file representative of the reference speech sample, the reference speech digital data file including information indicative of sound intensity levels as a function of time over a chronological duration of the reference speech digital data file;

in a computer process, generating a student speech digital data file representative of the student speech sample, the student speech digital data file including information indicative of sound intensity levels as a function of time over a chronological duration of the student speech digital data file;

in a computer process, segmenting the reference speech digital data file into a reference speech-unit data file the reference speech-unit data file including a sequence of speech units over a reference speech sample duration;

in a computer process, segmenting the student speech digital data file into a student speech-unit data file, the student speech-unit data file including a sequence of speech units over a student speech sample duration;

in a computer process, generating a reference speech-unit duration disparity based on a duration disparity between a first speech unit and a second speech unit, that is not adjacent to said first speech unit, within the reference speech-unit data file;

in a computer process, generating a student speech-unit duration disparity based on a duration disparity between third and fourth speech units in the student speech-unit data file that correspond to locations in the sequence of speech units of the first and second speech units in the reference speech-unit data file;

in a computer process, calculating a difference between the student speech-unit duration disparity and the reference speech-unit duration disparity; and

in a computer process, generating a sequence of student speech-unit duration differences by calculating a difference in duration between each speech unit in a sequence of speech units in the student speech sample and each other speech unit in the student speech sample.

7. The method of claim 6, further comprising:

in a computer process, calculating a difference in duration between each speech unit in a sequence of speech units in the reference speech sample and each other speech unit in the student speech sample, thereby generating a sequence of reference speech-unit duration differences corresponding in length to the sequence of student speech-unit duration differences.

8. The method of claim 7, further comprising:

in a computer process, summing an absolute value of each of the difference between a student speech-unit duration difference from the sequence of student speech-unit duration differences and a respective reference speech-unit duration difference from the sequence of reference speech-unit duration differences, to provide a difference sum;

in a computer process, summing an absolute value of each of the difference between a student speech-unit duration difference from the sequence of student speech-unit duration differences and an absolute value of a respective reference speech-unit duration difference from the sequence of reference speech-unit duration differences;

to provide a total absolute sum; and

in a computer process, dividing the difference sum by the total absolute sum to provide the Pairwise variability error (PVE).

9. The method of claim 6, wherein a type of speech-unit is at least one of phonemes, syllables, or words.

10. The method of claim 6, further comprising:

scaling at least one of the reference speech sample or the student speech sample so that the total duration of the reference speech sample and the student speech sample are substantially the same.

11. The method of claim 6, further comprising:

in a computer process, calculating a difference in duration between each speech unit in a sequence of speech units in the reference speech sample and each other speech unit in the student speech sample, thereby generating a sequence of reference speech-unit duration differences corresponding in length to the sequence of student speech-unit duration differences.

determining the most granular level of speech in which a number of speech-units is the same for the reference speech sample and the student speech sample; and

comparing a characteristic of each speech-unit in the sequence of speech-units in the student speech sample with the characteristic of the corresponding speech-unit in the sequence of speech-units in the reference speech sample.

12. A method, comprising:

generating a reference speech-unit data file from a native language speaker sample and a student speech-unit data file from a language student speech sample, the reference speech-unit data file including a plurality of reference speech units from the native language speaker sample and the student speech-unit data file including a plurality of student speech units from the language student speech sample; and

in a computer process, generating a plurality of student speech-unit duration differences by calculating a difference in duration between each student speech unit from the plurality of student speech units and each other speech unit from the plurality of student speech units.

13. The method of claim 12, wherein a type of speech-unit is at least one of phonemes, syllables, or words.

14. The method of claim 12, further comprising:

scaling at least one of the native language speaker sample or the student speech sample so that the total duration of the native language speaker sample and the student speech sample are substantially the same.