Method and apparatus for adapting reference templates

1. A method comprising: adapting a reference template stored in a computer-readable memory by replacing only a portion of existing data within said reference template with a weighted interpolation of said portion and corresponding data within a test utterance, wherein the reference template includes data indicative of physical characteristics of a voice of a person.

2. The method of claim 1 , further comprising:

comparing reference templates to said test utterance; and

selecting for adaptation said reference template that is most similar to said test utterance.

3. The method of claim 1 further comprising:

storing said test utterance before adapting the reference template.

4. A method comprising: adapting two or more reference templates stored in a computer-readable memory by replacing only corresponding portions of existing data within said reference template with a weighted interpolation of said corresponding portions and corresponding data within a test utterance, wherein each template includes data indicative of physical characteristics of a voice of a person.

5. The method of claim 4 , wherein adapting said reference templates comprises:

performing a dynamic time warp alignment between said reference templates and said test utterance;

matching one or more frames of said test utterance to corresponding frames in each of said reference templates;

selecting one of said corresponding frames that is most similar to said one or more frames of said test utterance as a best-match frame; and

replacing original data within said best-match frame with a weighted interpolation of said original data and data within said one or more frames of said test utterance.

6. The method of claim 5 further comprising:

repeating the operations of matching, selecting and replacing for other frames of said test utterance.

7. The method of claim 4 further comprising:

storing said test utterance before adapting the reference templates.

8. A computer-readable storage medium having stored thereon instructions that, when executed by a computing platform, result in:

adapting a reference template by replacing only a portion of existing data within said reference template with a weighted interpolation of said portion and corresponding data within a test utterance, wherein the reference template includes data indicative of physical characteristics of a voice of a person.

9. The computer-readable storage medium of claim 8 , wherein the instructions when executed further result in:

comparing reference templates to said test utterance; and

selecting for adaptation said reference template that is most similar to said test utterance.

10. An apparatus comprising:

a memory to store one or more reference templates, wherein each reference template includes data indicative of physical characteristics of a voice of a person; and

a processing unit coupled to said memory, said processing unit to adapt one or more of said reference templates using a test utterance by replacing only a portion of existing data within at least one of said reference templates with a weighted interpolation of said portion and corresponding data within said test utterance.

11. The apparatus of claim 10 , wherein said processing unit is to replace original data within frames of more than one of said reference templates, with a weighted interpolation of said original data and data within one or more corresponding frames of said test utterance.

12. An apparatus comprising:

a crystal microphone to receive a test utterance;

a memory to store one or more reference templates, wherein each reference template includes data indicative of physical characteristics of a voice of a person; and

a processing unit coupled to said memory, said processing unit to adapt one or more of said reference templates by replacing at least part of existing data within at least one of said reference templates with a weighted interpolation of said existing data and corresponding data within a test utterance;

wherein adapting said one or more reference templates comprises:

performing a dynamic time warp alignment between one of said reference templates and said test utterance; and

replacing only a portion of original data within said one reference template with a weighted interpolation of said portion and data within one or more corresponding frames of said test utterance.

13. The apparatus of claim 12 wherein said processing unit is to perform a dynamic time warp alignment between said reference templates and said test utterance.

14. The apparatus of claim 12 , wherein said processing unit is to replace original data within frames of more than one of said reference templates, with a weighted interpolation of said original data and data within one or more corresponding frames of said test utterance.

15. A computer-readable storage medium having stored thereon instructions that, when executed by a computing platform, result in:

adapting two or more reference templates by replacing only corresponding portions of existing data within said reference templates with a weighted interpolation of said corresponding portions of existing data and corresponding data within a test utterance, wherein each reference template includes data indicative of physical characteristics of a voice of a person.

16. An apparatus comprising:

a memory to store a plurality of reference templates, wherein each reference template includes data indicative of physical characteristics of a voice of a person; and

a processing unit coupled to said memory, said processing unit to adapt at least two of said plurality of reference templates using a test utterance by replacing only corresponding portions of existing data within said at least two of said plurality of reference templates with a weighted interpolation of said corresponding portions of existing data and corresponding data within said test utterance.

17. An apparatus according to claim 16 , further comprising:

a crystal microphone to receive said test utterance.

18. The method of claim 1 , wherein adapting said reference template comprises:

performing a dynamic time warp alignment between said reference template and said test utterance; and

replacing original data within a frame of said reference template with a weighted interpolation of said original data within said frame and data within one or more corresponding frames of said test utterance.

19. The computer-readable storage medium of claim 8 , wherein adapting said reference template comprises:

performing a dynamic time warp alignment between said reference template and said test utterance; and

replacing original data within a frame of said reference template with a weighted interpolation of said original data within said frame and data within one or more corresponding frames of said test utterance.

20. The apparatus of claim 10 , wherein the processing unit is configured to:

perform a dynamic time warp alignment between one of said reference templates and said test utterance, and

replace original data within a frame of said one reference template with a weighted interpolation of said original data within said frame and data within one or more corresponding frames of said test utterance.