Recognition on voice-prints of vowels using neural network
Yoshio Kaji, Takahiro Moriyama, Tomoyuki USHIDA · 2002
The aim of study is the recognition on the voice-prints. This paper describes how raw vowels in signals consisting of voices and environmental noise are recognized using a neural network method for the binary two dimensional lattice array composed of 10/spl times/10=100 for the normalized DFT of voices.