Metraipolis › archive › dc_ca6dd85bc063

Domain-Adaptive Dual-Gating Mixture of Experts for Generalizable Speech Deepfake Detection — arXiv:2609.33709

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.33709
License (as archived): CC0-1.0
Archived: 2026-10-08T01:37:16.346Z
SHA-256 of the archived text: 50a593b5641f67dac33ae4a5b50d654805e6b8fbbbfaa26246bc6ab3bce28635
Stable URL of this page: https://metraipolis.com/documento/dc_ca6dd85bc063

How to cite

Domain-Adaptive Dual-Gating Mixture of Experts for Generalizable Speech Deepfake Detection — arXiv:2609.33709 — arXiv (en), https://arxiv.org/abs/2609.33709 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ca6dd85bc063 (Metraipolis archive dc_ca6dd85bc063, sha256 50a593b5641f67dac33ae4a5b50d654805e6b8fbbbfaa26246bc6ab3bce28635)

BibTeX

@misc{dcca6dd85bc063,
  title = {Domain-Adaptive Dual-Gating Mixture of Experts for Generalizable Speech Deepfake Detection — arXiv:2609.33709},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.33709},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ca6dd85bc063; sha256 50a593b5641f67dac33ae4a5b50d654805e6b8fbbbfaa26246bc6ab3bce28635},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.33709. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Recent advances in speech deepfake detection (SDD) have leveraged the Mixture of Experts (MoE) to enhance generalization capacity. However, existing gating networks often overlook the acoustic and temporal cues of deepfakes. In this work, we propose a novel domain-adaptive dual-gating MoE (DADGMoE) framework for SDD under unseen attack types and acoustic conditions. Our innovative dual-gating mechanism leverages Sinc-layer-based filters to process both low-level acoustic signals (raw waveforms) and high-level speech representations from a large self-supervised learning (SSL) model. It further incorporates domain prototypes to guide expert routing based on implicit deepfake patterns. The lightweight affine experts process the routed inputs. Experiments show that our DADGMoE significantly outperforms the baseline, achieving up to a 40.8% relative EER reduction on challenging out-of-dataset benchmarks. This framework demonstrates superior generalization capabilities and efficient design.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in eess.AS).
Categorie arXiv: Audio and Speech Processing (eess.AS) ; Sound (cs.SD).
Pagina del lavoro: https://arxiv.org/abs/2609.33709 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-08T01:11:22.062Z dalla pagina https://arxiv.org/catchup/eess.AS/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).