用专家路由精调语音伪造检测模型，省算力还更准。
MoLEx: Mixture of LoRA Experts in Speech Self-Supervised Models for Audio Deepfake Detection
- 用低秩适配+专家路由组合，只动部分参数实现高效微调。
- 在ASVSpoof 5数据集上达到5.56%的最低错误率（无数据增强）。
- 能动态切换专家应对新攻击，适合需要快速适配的场景。
尽管基于自监督学习（SSL）的模型显著提升了语音深度伪造检测的准确性，但全量微调成本高昂。为此，我们提出一种参数高效框架——混合低秩适配专家（MoLEx），将低秩适配（LoRA）与专家混合（MoE）路由器结合。该方法保留预训练SSL模型的知识，仅高效微调选定专家，显著降低训练开销并保持强鲁棒性。推理时观察到，路由器会为相似攻击重复激活相同专家，而对新型伪造则切换至其他专家，验证了其领域感知适应能力。此外，通过新增专家即可实现领域迁移，无需修改整个模型。我们在ASVSpoof 5数据集上进行主要评估，未使用数据增强即取得5.56%的最优等错误率（EER）。
原文摘要 · Abstract (English)
While self-supervised learning (SSL)-based models have boosted audio deepfake detection accuracy, fully finetuning them is computationally expensive. To address this, we propose a parameter-efficient framework that combines Low-Rank Adaptation with a Mixture-of-Experts router, called Mixture of LoRA Experts (MoLEx). It preserves pre-trained knowledge of SSL models while efficiently finetuning only selected experts, reducing training costs while maintaining robust performance. The observed utility of experts during inference shows the router reactivates the same experts for similar attacks but switches to other experts for novel spoofs, confirming MoLEx's domain-aware adaptability. MoLEx additionally offers flexibility for domain adaptation by allowing extra experts to be trained without modifying the entire model. We mainly evaluate our approach on the ASVSpoof 5 dataset and achieve the state-of-the-art (SOTA) equal error rate (EER) of 5.56% on the evaluation set without augmentation.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。