AI text-to-speech name-calling system for the 57th Congregation

During Congregation ceremonies, graduates’ names are traditionally read by a human announcer, which can be error-prone. For the upcoming Congregation, an AI name-reading system will replace the human announcer.

The AI name-reading system utilises a high-fidelity text-to-speech (TTS) engine tailored for the University’s 57th Congregation. Leveraging advanced neural voice models, it ensures each graduate’s name is announced with dignity, natural phrasing, and phonetic accuracy.

The system delivers consistent, high-quality pronunciation while avoiding the vocal fatigue and tonal variation that human readers may experience over a long event. Its multilingual capability, supported by advanced neural models, provides precise pronunciation in Cantonese, Putonghua, and English. It also supports customisation and fine-tuning through a review interface that allows designated staff to review, refine, and approve pronunciations, especially for uncommon names.

These features will introduce modern and intelligent elements to the traditional Congregation ceremony.