PulseAugur
EN
LIVE 04:15:35

New RAT training improves deepfake detection accuracy

Researchers have developed a new training strategy called Reference-Augmented Training (RAT) to improve the detection of audio deepfakes. This method, detailed in a new paper, surprisingly enhances deepfake detection even when the reference audio is absent or mismatched during inference. By employing RAT, the system achieves state-of-the-art performance on the ASVspoof 5 benchmark, outperforming larger ensemble systems with a single detector. AI

IMPACT Improves deepfake detection capabilities, potentially enhancing security against voice-based impersonation.

RANK_REASON The cluster contains an academic paper detailing a new method for AI safety research. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

COVERAGE [1]

  1. arXiv cs.AI TIER_1 English(EN) · Kamil Malinka ·

    RAT: Reference-Augmented Training for ASV Anti-Spoofing

    We introduce a spoofing countermeasure architecture conditioned on speaker-reference recordings, but observe that it converges to a solution that effectively ignores the reference during inference. Surprisingly, training with a reference channel induces invariance that improves d…