You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

WS2-audio — Unlabeled Speech Segments

Audio-only dataset of VAD-segmented speech clips. No transcripts. See AUDIO_DATASET_SCHEMA.md in the WorldSpeech repo for the full spec.

Downloads last month
-