The VoxBlink2 dataset is a Large Scale speaker recognition dataset with 100K+ speakers obtained from YouTube platform. This repository provides guidelines to build the corpus and relative resources to reproduce the results in our article . For more introduction, please see cite. If you find this repository helpful to your research, don't forget to give us star🌟.
Let's start with obtaining the resource files and decompressing… See the full description on the dataset page:
https://huggingface.co/datasets/saeedzou/ScriptsForVoxBlink2.
