Closed ersigcha23 closed 7 years ago
Read https://github.com/alumae/gst-kaldi-nnet2-online:
Full results can now inlude word alignment information for the best hypothesis. In order to activate this, set the word-boundary-file property to point to the word_boundary.int file typically found in the data/lang/phones/ subdirectory of a trained system.
Hi
I want to know how can I use the segmentation feature?
Thanks