Vitual Avatar Development System
Xubin Sun
Co-Presenters: Individual Presentation
College: The Dorothy and George Hennings College of Science, Mathematics and Technology
Major: Computer Science
Faculty Research Mentor: Malihe Aliasgari
Abstract:
This research explores the development of an interactive virtual avatar system equipped with real-time Text-to-Speech (TTS) and Speech-to-Text (STT) functionalities, designed for seamless two-sided communication. The system integrates Confi UI for a user-friendly interface and is deployed via Docker for efficient, cross-platform compatibility. The avatar is designed with dynamic facial expressions and lip-syncing capabilities to enhance user engagement during interactions. A secure web-based platform with user authentication ensures privacy and controlled access. The project incorporates multi-language support and customizable voice options through API integrations, enhancing accessibility and personalization.A key focus of this research is on ensuring system scalability, security, and performance optimization. The containerized architecture, built on Docker, allows for easy deployment and maintenance while supporting multiple concurrent users. Security protocols, including data encryption and GDPR-compliant user authentication, safeguard user information. The project also emphasizes system reliability through comprehensive testing strategies, including functional, performance, and usability testing. The findings from this research contribute to advancing virtual agent technology, enabling more natural and secure human-computer interactions across diverse applications.