Open ghost opened 7 years ago
I have a set of labelled data for my audio files e.g.
START, STOP, DUR, PHON 0.00, 0.21, 21, pau 0.21, 0.27, 6, b 0.27, 0.31, 4, r 0.31, 0.44, 11, ih0 0.44, 0.52, 8, t 0.52, 0.61, 9, ae1 0.61, 0.70, 9, n ....
Is it possible to use this hand labelled data instead of automatically aligned data? I want to improve the accuracy of phoneme segmentation.
Thank you,
I have a set of labelled data for my audio files e.g.
Is it possible to use this hand labelled data instead of automatically aligned data? I want to improve the accuracy of phoneme segmentation.
Thank you,