experiments/STT
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Merge pull request #2200 from rhamnett/patch-2

Browse Source 
Allow for different sample rate exports
This commit is contained in:
Reuben Morais 2019-07-04 15:04:22 +00:00 committed by GitHub
commit 5c8af86aca
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 5 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
bin/import_librivox.py
View File

@ -19,6 +19,8 @@ from sox import Transformer
from util.downloader import maybe_download from util.downloader import maybe_download
from tensorflow.python.platform import gfile from tensorflow.python.platform import gfile
SAMPLE_RATE = 16000
def _download_and_preprocess_data(data_dir): def _download_and_preprocess_data(data_dir):
# Conditionally download data to data_dir # Conditionally download data to data_dir
print("Downloading Librivox data set (55GB) into {} if not already present...".format(data_dir)) print("Downloading Librivox data set (55GB) into {} if not already present...".format(data_dir))
@ -168,7 +170,9 @@ def _convert_audio_and_split_sentences(extracted_dir, data_set, dest_dir):
flac_file = os.path.join(root, seqid + ".flac") flac_file = os.path.join(root, seqid + ".flac")
wav_file = os.path.join(target_dir, seqid + ".wav") wav_file = os.path.join(target_dir, seqid + ".wav")
if not os.path.exists(wav_file): if not os.path.exists(wav_file):
Transformer().build(flac_file, wav_file) tfm = Transformer()
tfm.set_output_format(rate=SAMPLE_RATE)
tfm.build(flac_file, wav_file)
wav_filesize = os.path.getsize(wav_file) wav_filesize = os.path.getsize(wav_file)
files.append((os.path.abspath(wav_file), wav_filesize, transcript)) files.append((os.path.abspath(wav_file), wav_filesize, transcript))