This repository hosts the core materials and code developed for the Bachelor’s thesis entitled "Design and Development of an AI-based Virtual Teacher for Education", completed by the author in 2025 as part of a Bachelor’s Degree in Computer and Software Engineering.
MGVT is an application that adopts the architectural principles of the MARE framework to implement an advanced AI-driven Virtual Teacher. It combines natural language processing, real-time speech synthesis, and 3D avatar rendering in a modular design.
This project also led to the creation of several specialized components, including:
- Live functional demo: A Unity3D application featuring real-time lipsync and a distributed asynchronous pipeline.
- VerbaManent: A customized fork of Coqui-TTS, restructured for improved modularity, phoneme alignment, and performance on consumer hardware.
- MARE framework: A modular evolution of the original RAEF architecture described in the author's thesis, built on virtual roles (Writer, Reviewer, Dubber, Courier) to optimize text-to-speech generation in educational scenarios.
This project is released under the GNU General Public License v3 (GPL-3.0),
to promote open-source and academic research use.
Any proprietary, closed-source, or commercial use requires prior written authorization.
See the LICENSE file for complete terms.
For commercial licensing or special agreements, please contact Michele Giordano.
- VerbaManent
- MARE framework
- Upcoming publications will further describe the MGVT framework and its potential applications in education and beyond.
For questions or collaboration opportunities, please contact:
giordano.michele@gmail.com