Cross-lingual Voice Conversion

Author: Anderson Fraiha Machado

Advisor: Marcelo Queiroz

Voice Conversion is defined as the process of acoustical features conversion that sentences uttered by a source speaker are transformed into sentences that sound as sentences uttered by the target speaker.

The interest in solve this problem is that it can be applied in:

  • Voice imitative: the personification of Interactive Systems Text-To-Speech (TTS) and virtual performers, development of karaoke systems that emulate the voice of the original singer and Playback voice of celebrities who have died.
  • Multimedia applications, educational tools for foreign language learning, virtual custom interpreters. Other applications: development of systems that circumvent Biometric authentication systems and development of voice systems for voice restoration.

The objective of this project is inspection and evaluation the “quality” of the tools used in voice conversion systems in the literature to create a framework to perform a conversion of high fidelity.

This project was introduced in this research group from the Anderson Fraiha Machado, in his PhD course, and developed a little system in MATLAB which can be accessed here. His thesis (in Portuguese) can be accessed here.

At present we have four published works:

The project is a startpoint to developing of systems of speech processing, and any help is very welcome …

;-)