A user friendly interface operated by the improved DTW method

Author

Chu, MYung-Kyung ; Sohn, Young-sun

Author_Institution

Dept. of Inf./Commun. Eng., Tongmyong Univ. of Inf. Technol., Pusan, South Korea

Volume

3

fYear

2001

fDate

6/23/1905 12:00:00 AM

Firstpage

1187

Lastpage

1190

Abstract

This paper introduces an interface that moves its cursor by vocal commands instead of the mouse movement through fuzzy inference. Due to the short length of the vocal orders, the DTW method is applied to recognize each separate word. However, if the length of the commands is similar, the DTW method will choose a command with the least difference among the standard patterns. For example, if the command ´move extremely´ is inputted, then the command ´extremely right´ which is similar in length would be applied in some cases. To solve this type of error, we inferred the value of DTW error distance and the critical value to the voice length of standard patterns using fuzzy inference that leads to a command. for acceptance or rejection. If the case is ambiguous, then the system asks the user to make a decision

Keywords

fuzzy set theory; inference mechanisms; linear predictive coding; speech recognition; speech-based user interfaces; DTW method; critical value; error distance; fuzzy inference; linear prediction coding; mouse movement; speech recognition; user friendly interface; vocal commands; vocal orders; voice recognition; Discrete wavelet transforms; Fuzzy systems; Humans; Inference algorithms; Information technology; Man machine systems; Mice; Natural languages;

fLanguage

English

Publisher

ieee

Conference_Titel

Fuzzy Systems, 2001. The 10th IEEE International Conference on

Conference_Location

Melbourne, Vic.

Print_ISBN

0-7803-7293-X

Type

conf

DOI

10.1109/FUZZ.2001.1008868

Filename

1008868