Multilingual Dataset Integration Strategies for Robust Audio Deepfake Detection: A SAFE Challenge System

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Ali, Hashim, Subramani, Surya, Bollinani, Lekha, Adupa, Nithin Sai, El-Loh, Sali, Malik, Hafiz
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918155372199936
author Ali, Hashim
Subramani, Surya
Bollinani, Lekha
Adupa, Nithin Sai
El-Loh, Sali
Malik, Hafiz
author_facet Ali, Hashim
Subramani, Surya
Bollinani, Lekha
Adupa, Nithin Sai
El-Loh, Sali
Malik, Hafiz
contents The SAFE Challenge evaluates synthetic speech detection across three tasks: unmodified audio, processed audio with compression artifacts, and laundered audio designed to evade detection. We systematically explore self-supervised learning (SSL) front-ends, training data compositions, and audio length configurations for robust deepfake detection. Our AASIST-based approach incorporates WavLM large frontend with RawBoost augmentation, trained on a multilingual dataset of 256,600 samples spanning 9 languages and over 70 TTS systems from CodecFake, MLAAD v5, SpoofCeleb, Famous Figures, and MAILABS. Through extensive experimentation with different SSL front-ends, three training data versions, and two audio lengths, we achieved second place in both Task 1 (unmodified audio detection) and Task 3 (laundered audio detection), demonstrating strong generalization and robustness.
format Preprint
id arxiv_https___arxiv_org_abs_2508_20983
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Multilingual Dataset Integration Strategies for Robust Audio Deepfake Detection: A SAFE Challenge System
Ali, Hashim
Subramani, Surya
Bollinani, Lekha
Adupa, Nithin Sai
El-Loh, Sali
Malik, Hafiz
Audio and Speech Processing
Machine Learning
The SAFE Challenge evaluates synthetic speech detection across three tasks: unmodified audio, processed audio with compression artifacts, and laundered audio designed to evade detection. We systematically explore self-supervised learning (SSL) front-ends, training data compositions, and audio length configurations for robust deepfake detection. Our AASIST-based approach incorporates WavLM large frontend with RawBoost augmentation, trained on a multilingual dataset of 256,600 samples spanning 9 languages and over 70 TTS systems from CodecFake, MLAAD v5, SpoofCeleb, Famous Figures, and MAILABS. Through extensive experimentation with different SSL front-ends, three training data versions, and two audio lengths, we achieved second place in both Task 1 (unmodified audio detection) and Task 3 (laundered audio detection), demonstrating strong generalization and robustness.
title Multilingual Dataset Integration Strategies for Robust Audio Deepfake Detection: A SAFE Challenge System
topic Audio and Speech Processing
Machine Learning
url https://arxiv.org/abs/2508.20983