1 paper · 1 filter
Sung-Lin Yeh, Yen Meng, Hao Tang
There is an increasing interest in obtaining accurate word-level timestamps from strong automatic speech recognizers, in particular Whisper. Existing approaches either require addi…