README.md

metadata

title: VoxCeleb2 Dataset
license: mit
dataset_info:
  features:
    - name: audio
      dtype: audio
    - name: speaker_id
      dtype: string
    - name: video_id
      dtype: string
task_categories:
  - automatic-speech-recognition
  - speaker-verification
  - audio-classification
language:
  - en
pretty_name: VoxCeleb2 Dataset
size_categories:
  - 100K<n<1M

VoxCeleb2 Dataset

This is the VoxCeleb2 dataset, a large-scale speaker identification dataset.

Dataset Description

VoxCeleb2 contains over 1 million utterances for 6,112 celebrities, extracted from videos uploaded to YouTube.

Files

• vox2_dev_mp4_part*: Multipart archive containing MP4 video files
• vox2_dev_txt: Text files with speaker/utterance metadata
• vox2_meta.csv: Dataset metadata

Usage

To extract the multipart archive:

# Using 7zip
7z x vox2_dev_mp4_partaa

# Or using cat (Linux/Mac)
cat vox2_dev_mp4_parta* > combined.tar
tar -xf combined.tar

Citation

If you use this dataset, please cite:

@inproceedings{chung2018voxceleb2,
  title={VoxCeleb2: Deep Speaker Recognition},
  author={Chung, Joon Son and Nagrani, Arsha and Zisserman, Andrew},
  booktitle={INTERSPEECH},
  year={2018}
}

License

Please refer to the original VoxCeleb2 license terms.