search

gooofy / zamia-speech

Open tools and data for cloudless automatic speech recognition
GNU Lesser General Public License v3.0
444 stars 84 forks source link

G2P generates differnet phones than are in the dictionary #30

Closed dpny518 closed 6 years ago

dpny518 commented 6 years ago

I used the G2P to generate pronunciation for OOV, but the symbols its generate are not in the original dictionary

gooofy commented 6 years ago

take a look at https://github.com/gooofy/py-nltools/blob/master/nltools/phonetics.py here you will find converters between the differen phoneme alphabets we're using

dpny518 commented 5 years ago

I noticed that g2p produces a space between ' for example time t ' aI m what w ' V t you j ' u when it should be time t 'aI m what w 'V t you j 'u Use this to fix sed -i "s/' /'/g" generated_lexicon.txt