We present a dataset of audio deepfakes (and corresponding benign audio) for a set of politicians and other public figures, collected from publicly available sources such as social networks and video streaming platforms. For n = 58 celebrities and politicians, we collect both bona-fide and spoofed audio. In total, we collect 20.8 hours of bona-fide and 17.2 hours of spoofed audio. On average, there are 23 minutes of bona-fide and 18 minutes of spoofed audio per speaker.
The dataset is intended to be used for evaluating deepfake detection and voice anti-spoofing machine-learning models. It is especially useful to judge a model's capability to generalize to realistic, in-the-wild audio samples. Find more information in our paper, and download the dataset here.
The most interesting deepfake detection models we used in our experiments are open-source on GitHub:
This dataset and the associated documentation are licensed under the Apache License, Version 2.0.
@article{muller2022does,
title={Does audio deepfake detection generalize?},
author={M{\"u}ller, Nicolas M and Czempin, Pavel and Dieckmann, Franziska and Froghyar, Adam and B{\"o}ttinger, Konstantin},
journal={Interspeech},
year={2022}
}