ronggong / stutter-talk-preprocessing

Preprocessing code for stutter-talk "read as you type" app

stutter-talk-preprocessing

Preprocessing code for stutter-talk "read as you type" app

LibriSpeech tools

• audio_format_convert.sh convert audio from .flac to .wav, split transcription to line level lab.
• male_speaker_id.py extract the speaker ids from SPEAKER.txt.
• trim_audio.py trim utterance into word level samples using textgrid aligned.
• save_mongo_db.py insert word level samples to mongoDB.