Method and system for segmenting phonemes from voice signals

• US 20070150277A1
• Filed: 12/28/2006
• Published: 06/28/2007
• Est. Priority Date: 12/28/2005
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Kim, Hyun-Soo
• Granted Patent: US 8,849,662 B2
• Time in Patent Office: Days
• US Class Current: 704/254
• CPC Class Codes: G10L 15/04 Segmentation; Word boundary detection
• CPC Class Codes: G10L 2015/025 Phonemes, fenemes or fenones being the recognition units

Abstract

A method and a system for segmenting phonemes from voice signals. A method for accurately segmenting phonemes, in which a histogram showing a peak distribution corresponding to an order is formed by using a high order concept, and a boundary indicating a starting point and an ending point of each phoneme is determined by calculating a peak statistic based on the histogram. The phoneme segmentation method can remarkably reduce an amount of calculation, and has an advantage of being applied to sound signal systems which perform sound coding, sound recognition, sound synthesizing, sound reinforcement, etc.

Claims

1. A method for segmenting phonemes from voice signals, comprising:

extracting peak information from input voice signals;

calculating peak statistics per each frame based on the extracted peak information;

determining maxima among the calculated peak statistics per each frame; and

determining a valley between the maxima as a boundary between phonemes to perform a phoneme segmentation.

2. The method as claimed in claim 1, wherein the step of calculating the peak statistics per each frame based on the extracted peak information, comprises:

determining a length of a frame for calculating the peak statistics based on the extracted peak information;

forming a histogram showing a distribution of the extracted peak information with respect to the determined frame length; and

calculating the peak statistics using the histogram.

3. The method as claimed in claim 1, wherein the step of calculating the peak statistics per each frame based on the extracted peak information, comprises:

determining a peak order based on the extracted peak information;

determining a length of the frame for calculating the peak statistics when the peak order is determined;

forming a histogram showing a distribution of the extracted peak information with respect to the determined frame length; and

calculating the peak statistics using the histogram.

4. The method as claimed in claim 1, wherein the step of determining the maxima among the calculated peak statistics per each frame, comprises:

comparing the calculated peak statistics with a predetermined threshold in order to select the maxima, which are regarded as the phoneme, from the peak statistic; and

determining the maxima among the peak statistics greater than the threshold.

5. The method as claimed in claim 2, wherein the peak information is extracted in a state that a peak order is defined to a first order peak.

6. The method as claimed in claim 3, wherein the peak order is one of second peak order or third peak order.

7. The method as claimed in claim 3, wherein the length of the frame includes a 512 point frame as a default.

8. The method as claimed in claim 3, wherein the step of determining the peak order, comprises:

extracting peak information from voice signals on a time domain;

defining a peak order with respect to the extracted peak information;

comparing a peak measurement value of the defined peak order with a predetermined critical peak measurement value; and

determining a present peak order as the peak order when the peak measurement value is greater than the critical peak measurement value.

9. The method as claimed in claim 8, further comprising repeating the step of determining a present peak order unless the peak measurement value of the defined peak order is greater than the critical peak measurement value after the present peak order is allowed to increase, to define a novel peak order when the peak measurement value is below the critical peak measurement value, and a peak measurement value of the novel peak order is compared with the critical peak measurement value.

10. A system for segmenting phonemes from voice signals, comprising:

a peak information extractor for extracting peak information from input voice signals;

a peak statistic calculator for calculating peak statistics per each frame based on the extracted peak information; and

a boundary determination unit for determining maxima among the calculated peak statistics per each frame, and determining a valley between the maxima as a boundary between the phonemes in order to segment the phonemes.

11. The system as claimed in claim 10, further comprising:

a frame length determination unit for determining a length of a frame to calculate the peak statistics based on the extracted peak information; and

a histogram forming unit for forming a histogram showing a distribution of the extracted peak information with respect to the determined frame length.

12. The system as claimed in claim 11, wherein the peak statistic calculator calculates the peak statistics using the histogram.

13. The system as claimed in claim 10, wherein the boundary determination unit compares the peak statistics calculated by the peak statistic calculator with a predetermined threshold in order to select the maxima, which are regarded as the phonemes, among the peak statistics, and then determines the maxima among the greater peak statistics than the threshold.

14. The system as claimed in claim 11, wherein the frame includes a frame having a length of 512 points as a default.

15. The system as claimed in claim 11, further comprising a peak order determination unit for determining a peak order based on the extracted peak information.

16. The system as claimed in claim 14, wherein the peak order is one of second order or third order.

17. The system as claimed in claim 15, wherein the peak order determination unit defines a peak order with respect to the extracted peak information when the peak information is extracted by the peak information extractor from voice signals on the time domain, compares the peak measurement value of a present defined peak order with the predetermined critical peak measurement value, and determines a present peak order as the peak order when the peak measurement value is greater than the critical peak measurement value.

18. The system as claimed in claim 17, wherein the peak order determination unit increases the present peak order so as to define a novel peak order when the peak measurement value is below the critical peak measurement value, and then compares a peak measurement value of the novel peak order with the critical peak measure value, to determine the peak order when the novel peak measurement value is smaller than the critical peak measurement value.