TRANSFORMATION OF FORMANTS FOR VOICE CONVERSION USING ARTIFICIAL NEURAL NETWORKS

In this paper we propose a scheme for developing a voice conversion sy stem that converts the speech signal uttered by a source speaker to a speech signal having the voice characteristics of the target speaker. In particular, we address the issue of transformation of the vocal tra ct system features from one speaker to another. Formants are used to r epresent the vocal tract system features and a formant vocoder is used for synthesis. The scheme consists of a formant analysis phase, follo wed by a learning phase in which the implicit formant transformation i s captured by a neural network. The transformed formants together with the pitch contour modified to suit the average pitch of the target sp eaker are used to synthesize speech with the desired vocal tract syste m characteristics.