search

yh1008 / speech-to-text

mixlingual speech recognition system; hybrid (GMM+NNet) model; Kaldi + Keras
http://llcao.net/cu-deeplearning17/project.html
70 stars 19 forks source link

Mal-formed spk2gender file #7

Closed yh1008 closed 7 years ago

yh1008 commented 7 years ago

$ utils/validate_data_dir.sh --no-feats data/test

Mal-formed spk2gender file

right now the utt2spk shows less data/train/utt2spk | head -2 01FA-UI01FAZ_0101_0004721_0007863 01FA 01FA-UI01FAZ_0101_0008686_0012571 01FA

and spk2gender looks like less data/train/spk2gender | head -2 01FA F 02FA F

I might need to make 01FA-UI01FAZ this entire string as speaker_id

yh1008 commented 7 years ago

it should be lower case f and m not F and M ....

yh1008 commented 7 years ago

also there is a gender encoded as x... I assigned it to be f