The controllable Cantonese text-to-personalized speech generation system was developed by the research team led by Professor Helen Meng and Professor Xixin Wu. It uses speech language intelligence technology to analyze and process text, and convert it into corresponding speech content. At the same time, we provide personalized service, including customized pronunciation dictionaries and voice, to better meet the requirements of users and target scenarios. It provides more distinctive and convenient services for audio content creation.