VGGSound

VGG-Sound is an audio-visual correspondent dataset consisting of short clips of audio sounds, extracted from videos uploaded to YouTube.

• Homepage: https://www.robots.ox.ac.uk/~vgg/data/vggsound/
• Paper: https://arxiv.org/abs/2004.14368
• Github: https://github.com/hche11/VGGSound

Analysis

• 310+ classes: VGG-Sound contains audios spanning a large number of challenging acoustic environments and noise characteristics of real applications.
• 200,000+ videos: All videos are captured "in the wild" with audio-visual correspondence in the sense that the sound source is visually evident.
• 550+ hours: VGG-Sound consists of both audio and video. Each segment is 10 seconds long.

Download

We provide a csv file. For each YouTube video, we provide YouTube URLs, time stamps, audio labels and train/test split. Each line in the csv file has columns defined by here.

# YouTube ID, start seconds, label, train/test split. 

And you can download VGGSound directly from this repository.

License

The VGG-Sound dataset is available to download for commercial/research purposes under a Creative Commons Attribution 4.0 International License. The copyright remains with the original owners of the video. A complete version of the license can be found here.

Citation

Please cite the following if you make use of the dataset.

@InProceedings{Chen20,
  author       = "Honglie Chen and Weidi Xie and Andrea Vedaldi and Andrew Zisserman",
  title        = "VGGSound: A Large-scale Audio-Visual Dataset",
  booktitle    = "International Conference on Acoustics, Speech, and Signal Processing (ICASSP)",
  year         = "2020",
}