AINeutralarXiv – CS AI · 7h ago6/10
🧠
What Do Deepfake Speech Detectors Actually Hear?
Researchers developed an explainability pipeline that reveals what deepfake speech detectors actually focus on when identifying synthetic audio. The study found that three leading WavLM-based detectors rely on fundamentally different cues—environmental artifacts, phoneme distortions, and spectral patterns—despite achieving similar accuracy levels, with findings validated through causal masking experiments.