WaveFake: A Data Set to Facilitate Audio Deepfake Detection

TL;DR

This paper introduces WaveFake, a new audio dataset from multiple generative models and languages, along with baseline detection models, to advance research in audio deepfake detection.

Contribution

It provides a comprehensive audio dataset, signal processing techniques, and baseline models to address the gap in audio deepfake detection research.

Findings

Dataset includes nine sample sets from five architectures in two languages

Baseline models demonstrate initial detection capabilities

Provides foundational tools for future audio deepfake research

Abstract

Deep generative modeling has the potential to cause significant harm to society. Recognizing this threat, a magnitude of research into detecting so-called "Deepfakes" has emerged. This research most often focuses on the image domain, while studies exploring generated audio signals have, so-far, been neglected. In this paper we make three key contributions to narrow this gap. First, we provide researchers with an introduction to common signal processing techniques used for analyzing audio signals. Second, we present a novel data set, for which we collected nine sample sets from five different network architectures, spanning two languages. Finally, we supply practitioners with two baseline models, adopted from the signal processing community, to facilitate further research in this area.