speech_recognition/FunASR
1
0
Fork 0
mirror of https://github.com/modelscope/FunASR synced 2025-09-15 14:48:36 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

update docs

Browse Source
This commit is contained in:
speech_asr 2023-02-14 14:39:45 +08:00
parent ca1a92af0e
commit e180abe4fd
1 changed files with 26 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
funasr/utils/wav_utils.py
View File

@ -287,3 +287,29 @@ def generate_data_list(data_dir, dataset, nj=100):
wav_path = os.path.join(split_dir, str(i + 1), "wav.scp")
text_path = os.path.join(split_dir, str(i + 1), "text")
f_data.write(wav_path + " " + text_path + "\n")
def filter_wav_text(data_dir, dataset):
wav_file = os.path.join(data_dir,dataset,"wav.scp")
text_file = os.path.join(data_dir, dataset, "text")
with open(wav_file) as f_wav, open(text_file) as f_text:
wav_lines = f_wav.readlines()
text_lines = f_text.readlines()
os.rename(wav_file, "{}.bak".format(wav_file))
os.rename(text_file, "{}.bak".format(text_file))
wav_dict = {}
for line in wav_lines:
sample_name, wav_path = line.strip().split()
wav_dict[sample_name] = wav_path
text_dict = {}
for line in text_lines:
sample_name, txt = line.strip().split(" ", 1)
text_dict[sample_name] = txt
filter_count = 0
with open(wav_file) as f_wav, open(text_file) as f_text:
for sample_name, wav_path in wav_dict.items():
if sample_name in text_dict.keys():
f_wav.write(sample_name + " " + wav_path + "\n")
f_text.write(sample_name + " " + text_dict[sample_name] + "\n")
else:
filter_count += 1
print("{}/{} samples in {} are filtered because of the mismatch between wav.scp and text".format(len(wav_lines), filter_count, dataset))