泛化性音频深度伪造检测中的低秩适配器专家混合
Mixture of Low-Rank Adapter Experts in Generalizable Audio Deepfake Detection
AI总结：
针对基础模型微调后难以泛化至新型音频伪造方法的问题，提出混合LoRA专家方法，通过在注意力层集成多个低秩适配器并利用路由机制激活专家，有效提升了域外泛化能力并显著降低等错误率。
摘要：
诸如Wav2Vec2等基础模型在语音任务（包括音频深度伪造检测）的表征学习方面表现出色。然而，在固定的一组真实和伪造音频片段上进行微调后，它们通常无法泛化到训练中未涵盖的新型深度伪造方法。为了解决这一问题，我们提出了一种混合LoRA专家（mixture-of-LoRA-experts）方法，将多个低秩适配器（LoRA）集成到模型的注意力层中。路由机制选择性地激活专门化的专家，增强了对不断演变的深度伪造攻击的适应性。实验结果表明，我们的方法在域内和域外场景下均优于标准微调，降低了相对基线模型的等错误率。值得注意的是，我们最佳的MoE-LoRA模型将平均域外EER从8.55%降至6.08%，证明了其在实现泛化性音频深度伪造检测方面的有效性。
英文摘要：
Foundation models such as Wav2Vec2 excel at representation learning in speech tasks, including audio deepfake detection. However, after being fine-tuned on a fixed set of bonafide and spoofed audio clips, they often fail to generalize to novel deepfake methods not represented in training. To address this, we propose a mixture-of-LoRA-experts approach that integrates multiple low-rank adapters (LoRA) into the model's attention layers. A routing mechanism selectively activates specialized experts, enhancing adaptability to evolving deepfake attacks. Experimental results show that our method outperforms standard fine-tuning in both in-domain and out-of-domain scenarios, reducing equal error rates relative to baseline models. Notably, our best MoE-LoRA model lowers the average out-of-domain EER from 8.55\% to 6.08\%, demonstrating its effectiveness in achieving generalizable audio deepfake detection.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
