Can Emotion Fool Anti-spoofing?

TL;DR

This paper introduces EmoSpoof-TTS, a new emotional speech dataset, and GEM, a gated ensemble model that enhances anti-spoofing robustness against emotionally expressive synthetic speech, revealing current models' vulnerabilities.

Contribution

It provides the EmoSpoof-TTS dataset and proposes GEM, a novel emotion-aware anti-spoofing model, addressing the gap in robustness against emotional synthetic speech.

Findings

Existing models struggle with emotional speech

GEM improves detection across all emotions

Emotional data impacts anti-spoofing performance

Abstract

Traditional anti-spoofing focuses on models and datasets built on synthetic speech with mostly neutral state, neglecting diverse emotional variations. As a result, their robustness against high-quality, emotionally expressive synthetic speech is uncertain. We address this by introducing EmoSpoof-TTS, a corpus of emotional text-to-speech samples. Our analysis shows existing anti-spoofing models struggle with emotional synthetic speech, exposing risks of emotion-targeted attacks. Even trained on emotional data, the models underperform due to limited focus on emotional aspect and show performance disparities across emotions. This highlights the need for emotion-focused anti-spoofing paradigm in both dataset and methodology. We propose GEM, a gated ensemble of emotion-specialized models with a speech emotion recognition gating network. GEM performs effectively across all emotions and neutral state, improving defenses against spoofing attacks. We release the EmoSpoof-TTS Dataset: https://emospoof-tts.github.io/Dataset/