CONNECT WITH US

Taiwan government releases Chinese-language voice data to support AI development

Chloe Liao, Taipei; Willis Ke, DIGITIMES Asia

In efforts to accelerate Chinese-language AI development in Taiwan, the Ministry of Science and Technology (MOST) will release 400 hours of data from its AI voice databank set up in collaboration with local radio stations, with the data to be licensed free of charge to related government, industry, academic and research sectors, according to ministry sources.

The science ministry has established the AI voice database in cooperation with two local radios, which have accumulated 2,000-3,000 hours of data.

The 400 hours of AI voice data to be released will be uploaded onto the Data Market Platform operated by the National Center for High Performance Computing (NCHC) under the National Applied Research Laboratories (NARLabs). More data will be released later.

To facilitate the use of the data by relevant sectors, a professor at the National Taiwan University of Science and Technology has been invited to annotate the data to help users save the personnel and time on data annotation.