|
|
|
@ -1,10 +1,42 @@ |
|
|
|
import os |
|
|
|
os.chdir(r'C:\Users\Aki\source\repos\acoustic_model\acoustic_model') |
|
|
|
import sys |
|
|
|
|
|
|
|
import defaultfiles as default |
|
|
|
sys.path.append(default.forced_alignment_module_dir) |
|
|
|
from forced_alignment import pyhtk, convert_phone_set, scripts |
|
|
|
|
|
|
|
wav_dir = r'c:\Users\Aki\source\repos\acoustic_model\reus-test' |
|
|
|
reus_dir = r'c:\Users\Aki\source\repos\acoustic_model\reus-test' |
|
|
|
wav_dir = reus_dir |
|
|
|
wav_files = ['reus1008-reus.wav', |
|
|
|
'reus1167-man.wav', |
|
|
|
'reus3768-mantsje.wav'] |
|
|
|
|
|
|
|
word = 'reus' |
|
|
|
pronunciation_ipa = ['rø:s', 'mɑn', 'mɑntsjə'] |
|
|
|
|
|
|
|
for wav_file in wav_files: |
|
|
|
file_lab = os.path.join(reus_dir, wav_file.replace('.wav', '.lab')) |
|
|
|
file_dic = os.path.join(reus_dir, wav_file.replace('.wav', '.dic')) |
|
|
|
file_txt = os.path.join(reus_dir, wav_file.replace('.wav', '.txt')) |
|
|
|
|
|
|
|
# output htk dict file |
|
|
|
with open(file_dic, 'w', encoding="utf-8") as f: |
|
|
|
for ipa in pronunciation_ipa: |
|
|
|
cgn = convert_phone_set.ipa2cgn([ipa.replace(':', 'ː')]) |
|
|
|
barbara = convert_phone_set.cgn2barbara(cgn) |
|
|
|
f.write(word.upper() + '\t' + barbara + '\n') |
|
|
|
|
|
|
|
# output htk label file. |
|
|
|
pyhtk._create_label_file(word, file_lab) |
|
|
|
|
|
|
|
scripts.run_command([ |
|
|
|
'HVite','-T', '1', |
|
|
|
'-a', |
|
|
|
'-C', default.config_hvite, |
|
|
|
'-H', default.acoustic_model, |
|
|
|
'-m', |
|
|
|
'-i', file_txt, |
|
|
|
#'-S', script_file, |
|
|
|
file_dic, default.phonelist_txt, os.path.join(wav_dir, wav_file) |
|
|
|
]) |
yemaozi88
3 years ago