SynHate: Detecting Hate Speech in Synthetic Deepfake Audio

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Ranjan, Rishabh, Pipariya, Kishan, Vatsa, Mayank, Singh, Richa
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913884710895616
author Ranjan, Rishabh
Pipariya, Kishan
Vatsa, Mayank
Singh, Richa
author_facet Ranjan, Rishabh
Pipariya, Kishan
Vatsa, Mayank
Singh, Richa
contents The rise of deepfake audio and hate speech, powered by advanced text-to-speech, threatens online safety. We present SynHate, the first multilingual dataset for detecting hate speech in synthetic audio, spanning 37 languages. SynHate uses a novel four-class scheme: Real-normal, Real-hate, Fake-normal, and Fake-hate. Built from MuTox and ADIMA datasets, it captures diverse hate speech patterns globally and in India. We evaluate five leading self-supervised models (Whisper-small/medium, XLS-R, AST, mHuBERT), finding notable performance differences by language, with Whisper-small performing best overall. Cross-dataset generalization remains a challenge. By releasing SynHate and baseline code, we aim to advance robust, culturally sensitive, and multilingual solutions against synthetic hate speech. The dataset is available at https://www.iab-rubric.org/resources.
format Preprint
id arxiv_https___arxiv_org_abs_2506_06772
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SynHate: Detecting Hate Speech in Synthetic Deepfake Audio
Ranjan, Rishabh
Pipariya, Kishan
Vatsa, Mayank
Singh, Richa
Sound
Audio and Speech Processing
The rise of deepfake audio and hate speech, powered by advanced text-to-speech, threatens online safety. We present SynHate, the first multilingual dataset for detecting hate speech in synthetic audio, spanning 37 languages. SynHate uses a novel four-class scheme: Real-normal, Real-hate, Fake-normal, and Fake-hate. Built from MuTox and ADIMA datasets, it captures diverse hate speech patterns globally and in India. We evaluate five leading self-supervised models (Whisper-small/medium, XLS-R, AST, mHuBERT), finding notable performance differences by language, with Whisper-small performing best overall. Cross-dataset generalization remains a challenge. By releasing SynHate and baseline code, we aim to advance robust, culturally sensitive, and multilingual solutions against synthetic hate speech. The dataset is available at https://www.iab-rubric.org/resources.
title SynHate: Detecting Hate Speech in Synthetic Deepfake Audio
topic Sound
Audio and Speech Processing
url https://arxiv.org/abs/2506.06772