Intrinsic normalization and extrinsic denormalization of formant data of vowels

TL;DR

This paper introduces a combined normalization and denormalization method for vowel formant data that reduces talker variability and improves vowel classification accuracy.

Contribution

It proposes a novel speaker-extrinsic re-scaling procedure that enhances existing normalization techniques for better vowel space representation.

Findings

Improved vowel classification accuracy over existing methods

Reduced talker-induced spread in vowel formant data

Effective combination of intrinsic and extrinsic normalization techniques

Abstract

Using a known speaker-intrinsic normalization procedure, formant data are scaled by the reciprocal of the geometric mean of the first three formant frequencies. This reduces the influence of the talker but results in a distorted vowel space. The proposed speaker-extrinsic procedure re-scales the normalized values by the mean formant values of vowels. When tested on the formant data of vowels published by Peterson and Barney, the combined approach leads to well separated clusters by reducing the spread due to talkers. The proposed procedure performs better than two top-ranked normalization procedures based on the accuracy of vowel classification as the objective measure.