Publication Details
Diffuse or Confuse: A Diffusion Deepfake Speech Dataset
Malinka Kamil, Mgr., Ph.D. (DITS FIT BUT)
Hanáček Petr, doc. Dr. Ing. (DITS FIT BUT)
deepfakes, deepfake speech, dataset, diffusion, detection
Advancements in artificial intelligence and machine learning have significantly improved synthetic speech generation. This paper explores diffusion models, a novel method for creating realistic synthetic speech. We create a diffusion dataset using available tools and pretrained models. Additionally, this study assesses the quality of diffusion-generated deepfakes versus non-diffusion ones and their potential threat to current deepfake detection systems. Findings indicate that the detection of diffusion-based deepfakes is generally comparable to non-diffusion deepfakes, with some variability based on detector architecture. Re-vocoding with diffusion vocoders shows minimal impact, and the overall speech quality is comparable to non-diffusion methods.
@INPROCEEDINGS{FITPUB13230,
author = "Anton Firc and Kamil Malinka and Petr Han\'{a}\v{c}ek",
title = "Diffuse or Confuse: A Diffusion Deepfake Speech Dataset",
pages = "1--7",
booktitle = "2024 International Conference of the Biometrics Special Interest Group (BIOSIG)",
year = 2024,
location = "Darmstadt, DE",
publisher = "GI - Group for computer science",
ISBN = "978-3-88579-749-4",
doi = "10.1109/BIOSIG61931.2024.10786752",
language = "english",
url = "https://www.fit.vut.cz/research/publication/13230"
}