no code implementations • 19 Jul 2017 • Chia-Hao Shen, Janet Y. Sung, Hung-Yi Lee
We train SA from one language (source language) and use it to extract the vector representation of the audio segments of another language (target language).