sumansamui / utterance_to_words_splitting Goto Github PK
View Code? Open in Web Editor NEWThis project develops a new keyword database (containing audio files with a single word in it) from a standard speech corpus. So, the python scripts (in this repository) take a speech corpus (e.g. LibriSpeech) and decompose the original audio files (containing utterances) into audio files (each containing a single word).
License: Apache License 2.0