/Speech-Datasets-for-ASR

Download speech datasets (English and non-English) for Automatic Speech Recognition

Primary LanguageJupyter Notebook

Download Free Audio Data for (Automatic Speech Recognition) ASR

This repo contains some of the common publically available audio data that you can download for ASR or other speech activities.

Source Link Size (Hours)
Mozilla Common Voice 10.0-tr 79
Mozilla Common Voice 10.0-en 3050
Voxforge Voxforge-en-16kHz 130
Voxforge Voxforge-tr-16kHz 3

Download Speech Data for Other Languages

You can download and create manifest.jsonl for different languages supported by the dataset.

  • In the following script common_voice_dataset.ipynb change the version and language to you wish to download from Mozilla.
version = 'cv-corpus-8.0-2022-01-19'
language = 'cv-corpus-8.0-2022-01-19-en'
  • In the following script voxforge_dataset.ipynb change the language to you wish to download from Voxforge.
VOXFORGE_URL_16kHz = 'http://www.repository.voxforge1.org/downloads/en/Trunk/Audio/Main/16kHz_16bit/'

Reference Github Link: ASR-Audio-Data-Links