Third International Conference on Spoken Language Processing (ICSLP 94)

Yokohama, Japan
September 18-22, 1994

Phoneme-Level Voice Individuality Used in Speaker Recognition

Sadaoki Furui, Tomoko Matsui

NTT Human Interface Laboratories, Tokyo, Japan

This paper reports our recent investigation on voice individuality (speaker characterization) at the phoneme level of speech. The first half of the paper reports our perceptual experiments on speaker characterizing information included in continuous speech. Experimental results show that speaker-characterizing vowels differ from speaker to speaker. The latter half of the paper introduces a text-prompted speaker recognition method that we have recently proposed as an automatic speaker recognition method using phoneme units. In this method, the recognition system prompts each user with a new key sentence every time the system is used, and accepts the input utterance only when it decides that the registered speaker has uttered the prompted sentence.

Full Paper

Bibliographic reference.  Furui, Sadaoki / Matsui, Tomoko (1994): "Phoneme-level voice individuality used in speaker recognition", In ICSLP-1994, 1463-1466.