DocumentCode
1643189
Title
Speaker-independent vowel classification from inaccurate formant features
Author
Scalkwyk, Johan ; Vermeulen, Pieter ; Barnard, Etienne
Author_Institution
Pretoria Univ., South Africa
fYear
1992
fDate
9/11/1992 12:00:00 AM
Firstpage
55
Lastpage
57
Abstract
Formant extraction is a notoriously unreliable procedure. Neural networks on the other hand are able to deal with such inaccurate data. It is shown that a multilayer perceptron is able to classify five types of vowels with acceptable accuracy (approximately 74%) when operating on very simple formant-based features
Keywords
feature extraction; feedforward neural nets; speech recognition; accuracy; inaccurate formant features; multilayer perceptron; speaker-independent vowel classification; speech recognition; Autocorrelation; Data mining; Face recognition; Frequency; Humans; Linear predictive coding; Multi-layer neural network; Neural networks; Speech recognition; Transfer functions;
fLanguage
English
Publisher
ieee
Conference_Titel
Communications and Signal Processing, 1992. COMSIG '92., Proceedings of the 1992 South African Symposium on
Conference_Location
Cape Town
Print_ISBN
0-7803-0807-7
Type
conf
DOI
10.1109/COMSIG.1992.274313
Filename
274313
Link To Document