Result Details

SCDF: A Speaker Characteristics DeepFake Speech Dataset for Bias Analysis

STANĚK, V.; SRNA, K.; FIRC, A.; MALINKA, K. SCDF: A Speaker Characteristics DeepFake Speech Dataset for Bias Analysis. In Proceedings of the 24th International Conference of the Biometrics Special Interest Group (BIOSIG 2025). GI-Edition. Proceedings. Darmstadt: Gesellschaft für Informatik e.V., 2025. no. 367, p. 55-64.
Type
conference paper
Language
English
Authors
Staněk Vojtěch, Ing., DITS (FIT), DCGM (FIT)
Srna Karel, Bc., FIT (FIT), DCGM (FIT)
Firc Anton, Ing., Ph.D., DITS (FIT)
Malinka Kamil, doc. Mgr., Ph.D., DITS (FIT)
Abstract

Despite growing attention to deepfake speech detection, the aspects of
bias and fairness remain underexplored in the speech domain. To address
this gap, we introduce the Speaker Characteristics Deepfake (SCDF)
dataset: a novel, richly annotated resource enabling systematic
evaluation of demographic biases in deepfake speech detection. SCDF
contains over 237,000 utterances in a balanced representation of both
male and female speakers spanning five languages and a wide age range.
We evaluate several state-of-the-art detectors and show that speaker
characteristics significantly influence detection performance, revealing
disparities across sex, language, age, and synthesizer type. These
findings highlight the need for bias-aware development and provide a
foundation for building non-discriminatory deepfake detection systems
aligned with ethical and regulatory standards.

Keywords

Bias, Fairness, Dataset, Deepfake Speech, Anti-spoofing

URL
Published
2025
Pages
55–64
Journal
GI-Edition. Proceedings, vol. 2025, no. 367, ISSN
Proceedings
Proceedings of the 24th International Conference of the Biometrics Special Interest Group (BIOSIG 2025)
Conference
24th International Conference of the Biometrics Special Interest Group
Publisher
Gesellschaft für Informatik e.V.
Place
Darmstadt
DOI
BibTeX
@inproceedings{BUT198600,
  author="Vojtěch {Staněk} and Karel {Srna} and Anton {Firc} and Kamil {Malinka}",
  title="SCDF: A Speaker Characteristics DeepFake Speech Dataset for Bias Analysis",
  booktitle="Proceedings of the 24th International Conference of the Biometrics Special Interest Group (BIOSIG 2025)",
  year="2025",
  journal="GI-Edition. Proceedings",
  volume="2025",
  number="367",
  pages="55--64",
  publisher="Gesellschaft für Informatik e.V.",
  address="Darmstadt",
  doi="10.18420/biosig\{_}2025\{_}005",
  issn="1617-5468",
  url="https://doi.org/10.18420/biosig_2025_005"
}
Projects
Reliable, Secure, and Intelligent Computer Systems, BUT, Vnitřní projekty VUT, FIT-S-23-8151, start: 2023-03-01, end: 2026-02-28, completed
Research groups
IT Security Research Group (RG Security@FIT)
Departments
Back to top