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

Merge pull request #1647 from nicolaspanel/master

Browse Source 
[TrainingSpeech importer] generate english-complatible dataset
This commit is contained in:
lissyx 2018-10-15 16:00:58 +02:00 committed by GitHub
commit 5118f7ffed
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 20 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
bin/import_ts.py
View File

@ -3,6 +3,7 @@ from __future__ import absolute_import, division, print_function
# Make sure we can import stuff from util/
# This script needs to be run from the root of the DeepSpeech repository
import argparse
import os
import re
import sys
@ -11,6 +12,7 @@ import sys
sys.path.insert(1, os.path.join(sys.path[0], '..'))
import csv
import unidecode
import zipfile
from os import path
@ -25,7 +27,7 @@ ARCHIVE_DIR_NAME = 'ts_' + ARCHIVE_NAME
ARCHIVE_URL = 'https://s3.eu-west-3.amazonaws.com/audiocorp/releases/' + ARCHIVE_NAME + '.zip'
def _download_and_preprocess_data(target_dir):
def _download_and_preprocess_data(target_dir, english_compatible=False):
# Making path absolute
target_dir = path.abspath(target_dir)
# Conditionally download data
@ -33,7 +35,8 @@ def _download_and_preprocess_data(target_dir):
# Conditionally extract archive data
_maybe_extract(target_dir, ARCHIVE_DIR_NAME, archive_path)
# Conditionally convert TrainingSpeech data to DeepSpeech CSVs and wav
_maybe_convert_sets(target_dir, ARCHIVE_DIR_NAME)
_maybe_convert_sets(target_dir, ARCHIVE_DIR_NAME, english_compatible=english_compatible)
def _maybe_extract(target_dir, extracted_data, archive_path):
# If target_dir/extracted_data does not exist, extract archive in target_dir
@ -47,7 +50,8 @@ def _maybe_extract(target_dir, extracted_data, archive_path):
else:
print('Found directory "%s" - not extracting it from archive.' % archive_path)
def _maybe_convert_sets(target_dir, extracted_data):
def _maybe_convert_sets(target_dir, extracted_data, english_compatible=False):
extracted_dir = path.join(target_dir, extracted_data)
# override existing CSV with normalized one
target_csv_template = os.path.join(target_dir, 'ts_' + ARCHIVE_NAME + '_{}.csv')
@ -70,7 +74,7 @@ def _maybe_convert_sets(target_dir, extracted_data):
test_writer.writeheader()
for i, item in enumerate(data):
transcript = validate_label(cleanup_transcript(item['text']))
transcript = validate_label(cleanup_transcript(item['text'], english_compatible=english_compatible))
if not transcript:
continue
wav_filename = os.path.join(target_dir, extracted_data, item['path'])
@ -92,12 +96,22 @@ PUNCTUATIONS_REG = re.compile(r"[°\-,;!?.()\[\]*…—]")
MULTIPLE_SPACES_REG = re.compile(r'\s{2,}')
def cleanup_transcript(text):
def cleanup_transcript(text, english_compatible=False):
text = text.replace('’', "'").replace('\u00A0', ' ')
text = PUNCTUATIONS_REG.sub(' ', text)
text = MULTIPLE_SPACES_REG.sub(' ', text)
if english_compatible:
text = unidecode.unidecode(text)
return text.strip().lower()
def handle_args():
parser = argparse.ArgumentParser(description='Importer for TrainingSpeech dataset.')
parser.add_argument(dest='target_dir')
parser.add_argument('--english-compatible', action='store_true', dest='english_compatible', help='Remove diactrics and other non-ascii chars.')
return parser.parse_args()
if __name__ == "__main__":
_download_and_preprocess_data(sys.argv[1])
cli_args = handle_args()
_download_and_preprocess_data(cli_args.target_dir, cli_args.english_compatible)