Hi,
I am trying to use the HTS demo for speech synthesis. For training the
data, I created .lab files (after creating.utt files with the help of
.segment, .phrase,.syllable,.target,IntEvent files) and then tested with
the test files (HTS-demo_CMU-ARCTIC-SLT/data/labels/gen/*.lab ) and it
worked fine. I have a couple of queries.
1. How to generate the test .lab files if I use different set of test
sentences for generation?
Are /.segment, .phrase,.syllable,IntEvent/ files (without time
boundaries) only are sufficient to make .utt and then create .lab file
for generation of speech? or is there any executable/script to create
.lab file just by taking a test "sentence" as input.
2. I am trying understand the the
HTS-Demo_CMU-ARCTIC-SLT/data/questions/questions_qst001.hed file . I
think it is based on radio/darpa phone set. But how to create it if I
use different phone set with different test sentences (other than
provided in HTS -Demo)?
Can anyone please help me regarding this.
Thanks in Advance
--
Basha S M A