SEA-Spoof: Bridging The Gap in Multilingual Audio Deepfake Detection for South-East Asian

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Wu, Jinyang, Hou, Nana, Pan, Zihan, Zhang, Qiquan, Bhupendra, Sailor Hardik, Mondal, Soumik
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866908557968932864
author Wu, Jinyang
Hou, Nana
Pan, Zihan
Zhang, Qiquan
Bhupendra, Sailor Hardik
Mondal, Soumik
author_facet Wu, Jinyang
Hou, Nana
Pan, Zihan
Zhang, Qiquan
Bhupendra, Sailor Hardik
Mondal, Soumik
contents The rapid growth of the digital economy in South-East Asia (SEA) has amplified the risks of audio deepfakes, yet current datasets cover SEA languages only sparsely, leaving models poorly equipped to handle this critical region. This omission is critical: detection models trained on high-resource languages collapse when applied to SEA, due to mismatches in synthesis quality, language-specific characteristics, and data scarcity. To close this gap, we present SEA-Spoof, the first large-scale Audio Deepfake Detection (ADD) dataset especially for SEA languages. SEA-Spoof spans 300+ hours of paired real and spoof speech across Tamil, Hindi, Thai, Indonesian, Malay, and Vietnamese. Spoof samples are generated from a diverse mix of state-of-the-art open-source and commercial systems, capturing wide variability in style and fidelity. Benchmarking state-of-the-art detection models reveals severe cross-lingual degradation, but fine-tuning on SEA-Spoof dramatically restores performance across languages and synthesis sources. These results highlight the urgent need for SEA-focused research and establish SEA-Spoof as a foundation for developing robust, cross-lingual, and fraud-resilient detection systems.
format Preprint
id arxiv_https___arxiv_org_abs_2509_19865
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SEA-Spoof: Bridging The Gap in Multilingual Audio Deepfake Detection for South-East Asian
Wu, Jinyang
Hou, Nana
Pan, Zihan
Zhang, Qiquan
Bhupendra, Sailor Hardik
Mondal, Soumik
Sound
The rapid growth of the digital economy in South-East Asia (SEA) has amplified the risks of audio deepfakes, yet current datasets cover SEA languages only sparsely, leaving models poorly equipped to handle this critical region. This omission is critical: detection models trained on high-resource languages collapse when applied to SEA, due to mismatches in synthesis quality, language-specific characteristics, and data scarcity. To close this gap, we present SEA-Spoof, the first large-scale Audio Deepfake Detection (ADD) dataset especially for SEA languages. SEA-Spoof spans 300+ hours of paired real and spoof speech across Tamil, Hindi, Thai, Indonesian, Malay, and Vietnamese. Spoof samples are generated from a diverse mix of state-of-the-art open-source and commercial systems, capturing wide variability in style and fidelity. Benchmarking state-of-the-art detection models reveals severe cross-lingual degradation, but fine-tuning on SEA-Spoof dramatically restores performance across languages and synthesis sources. These results highlight the urgent need for SEA-focused research and establish SEA-Spoof as a foundation for developing robust, cross-lingual, and fraud-resilient detection systems.
title SEA-Spoof: Bridging The Gap in Multilingual Audio Deepfake Detection for South-East Asian
topic Sound
url https://arxiv.org/abs/2509.19865