DocumentCode :
2016502
Title :
Analysis of Algorithms for Phrase Recognition
Author :
Mirza, J.S. ; Umair, Muhammad ; Hayat, S.A. ; Hussain, Asif
Author_Institution :
Dept. of Comput. Sci., COMSATS Inst. of Inf. Technol., Lahore
fYear :
2005
fDate :
24-25 Dec. 2005
Firstpage :
1
Lastpage :
7
Abstract :
Three algorithms, designed to recognize vowels, are analyzed for their complexity class or growth pattern. First a databank was prepared from vowel loops of Peterson and Barney (1952) which was used for vowel recognition. The above three algorithms produced the same vowel-recognition rate of more than 78% on a given short phrase. This paper submits the result of our analysis of the three algorithms to investigate their growth pattern and decide which of the three is the fastest in making decision of vowel recognition. The simple system of vowel recognition, presented in this paper, which leads to phrasal recognition, can serve the purpose of automatic recognition of short phrases in a long wire-tapped message. The phrasal recognition can be advantageous to do a preliminary study whether or not a wire-tapped long message contained phrases of interest. If it does the intelligence agent can get primed to scrutinize the entire message. This scheme can spare a lot of time of heavily loaded intelligence agent deputed to do wire-tapping
Keywords :
computational complexity; speaker recognition; computational complexity; intelligence agent; long wire-tapped message; phrase recognition; vowel recognition; Algorithm design and analysis; Computer science; Educational institutions; Government; Intelligent agent; Pattern analysis; Pattern recognition; Speech analysis; Speech recognition; Testing;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
9th International Multitopic Conference, IEEE INMIC 2005
Conference_Location :
Karachi
Print_ISBN :
0-7803-9429-1
Electronic_ISBN :
0-7803-9430-5
Type :
conf
DOI :
10.1109/INMIC.2005.334417
Filename :
4133432
Link To Document :
بازگشت