Taiwanese Language AI TTS for Everyone.
A website where anyone can experience Taigi AI speech synthesis (TTS).
Experience nowTry these example sentences:
Demonstration of Taigi AI speech synthesis technology
Taigi is one of the most widely spoken languages in Taiwan. It is cherished and passed down through daily life, family and community conversations, and traditional culture.
Originating mainly from the Fujian region, Taigi has developed its own unique expressions and pronunciations over a long period in Taiwan. It is deeply rooted in Taiwan's rich culture, including songs, theater, and proverbs.
In recent years, efforts to learn and use Taigi have been expanding, and its preservation for future generations is attracting increasing attention.
Taigi speech synthesis presents several unique challenges::
This project has been developed using new technology approaches to overcome these challenges.
AI Researcher / Language Technology Developer
While studying law at university, I entered the world of programming through self-study. After graduation, I challenged myself with iOS app development and started working as a freelancer 11 years ago.
In 2014, I visited Taiwan for the first time and was deeply attracted by its culture, people, and language. I became fascinated by the depth and beauty of the Taigi language.
I want to connect the richness of the Taigi language to the future through the power of technology. With this passion, I launched the world's first Taigi AI Text-to-Speech (TTS) project and have been working on its development every day.
In this project, we have officially obtained permission to use the voice of Audrey Tang (唐鳳), a prominent Taigi figure, and have released a Taigi TTS utilizing her voice.
In 2014, I was fascinated by the beauty of the Taigi language during my visit to Taiwan. At the same time, I learned about its endangered status, which inspired me to start this project.Existing TTS technologies could not reproduce the complex phonetic system of Taigi, so it was necessary to develop a unique AI model.
After two years of research, we successfully generated natural Taigi speech by combining over 100 hours of audio data collected from various regions in Taiwan with new deep learning-based speech synthesis technology.
This project aims to be utilized in various fields such as education, entertainment, and cultural preservation.
info@taiwan-travel.net
東京都国立市東1丁目4−10 ホワイトパレスビル