SCDF: A Speaker Characteristics DeepFake Speech Dataset for Bias Analysis

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F26%3A0198600" target="_blank" >RIV/00216305:26230/26:0198600 - isvavai.cz</a>

Výsledek na webu

<a href="https://doi.org/10.18420/biosig_2025_005" target="_blank" >https://doi.org/10.18420/biosig_2025_005</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.18420/biosig_2025_005" target="_blank" >10.18420/biosig_2025_005</a>

Jazyk výsledku

angličtina

Název v původním jazyce

SCDF: A Speaker Characteristics DeepFake Speech Dataset for Bias Analysis

Popis výsledku v původním jazyce

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.

Název v anglickém jazyce

SCDF: A Speaker Characteristics DeepFake Speech Dataset for Bias Analysis

Popis výsledku anglicky

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.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2025

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Proceedings of the 24th International Conference of the Biometrics Special Interest Group (BIOSIG 2025)

ISBN

—

ISSN

1617-5468

e-ISSN

2944-7682

Počet stran výsledku

10

Strana od-do

55-64

Název nakladatele

Gesellschaft für Informatik e.V.

Místo vydání

Darmstadt

Místo konání akce

Darmstadt

Datum konání akce

24. 9. 2025

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—