EUI Students Develop AI Model That Speaks in Egyptian Dialect

A group of students from the Faculty of Computing and Information Sciences have developed a pioneering AI-powered text-to-speech model that speaks in the Egyptian dialect, marking a significant step in localizing artificial intelligence technologies for Arabic speakers.

Omar Samir, one of the students involved in the project, explained during an interview on Al Hayah Al Youm with host Lobna Assal that the team built a 17-hour audio dataset of spoken Egyptian Arabic. The audio was carefully segmented into short clips and stripped of background noise to help the AI model learn the dialect with precision.

The project not only showcases technical innovation but also opens doors for practical applications in entertainment and tourism. One team member noted that the technology could enhance the tourist experience by enabling interactive AI-driven guides at heritage sites, using the familiar Egyptian dialect to engage visitors.

The students are also exploring how AI can be used to promote tourism among youth, making it easier to organize youth-centric cultural events. Research is currently underway to study the role of young people in tourism marketing, to integrate AI tools to streamline event planning.

Beyond tourism, the model has strong potential in the education sector. The team highlighted its use in converting educational content into audio, particularly benefiting children and individuals with visual impairments. The AI tool could help visually impaired users access educational material in an intuitive and accessible way.

Notably, international companies have already expressed interest in the project, with requests for similar models tailored to other dialects and markets. The students announced that their Egyptian Arabic AI voice model is expected to be publicly available within a week.