Automatic vowel quality description using a variable mapping to an eight cardinal vowel reference set

This paper investigates the possibility of describing vowels phonetically using an automated method. Models of the phonetic dimensions of the vowel space are built using two multi-layer perceptrons trained using eight cardinal vowels. The paper aims to improve the positioning of vowels in the open-close dimension by experimenting with a parameter in the model α which is the parameter which controls the slope of the sigmoid function employed in the multi-layer perceptrons