用小参数量提升语音伪造检测精度，通过多尺度小波提示调优实现。
WaveSP-Net: Learnable Wavelet-Domain Sparse Prompt Tuning for Speech Deepfake Detection
- 融合小波变换与提示调优，构建可学习的前端特征提取器。
- 在Deepfake-Eval-2024和SpoofCeleb上超越多个顶尖模型。
- 适合需要轻量化高精度语音伪造检测的场景。
当前语音伪造检测的前端设计依赖于XLSR等大模型的全量微调，但该方法参数效率低，且在真实场景数据上泛化能力不足。为此，本文提出一类新型参数高效的前端架构，将提示调优与经典信号处理变换结合，包括基于傅里叶变换的FourierPT-XLSR，以及基于小波变换的WSPT-XLSR和Partial-WSPT-XLSR。进一步提出WaveSP-Net，由Partial-WSPT-XLSR前端与双向Mamba后端构成。该结构将多分辨率特征注入提示嵌入，增强对细微合成痕迹的定位能力，同时保持XLSR参数冻结。实验表明，WaveSP-Net在两个新且具有挑战性的基准测试集Deepfake-Eval-2024和SpoofCeleb上优于多个先进模型，仅使用少量可训练参数即获得显著性能提升。代码与模型已公开于https://github.com/xxuan-acoustics/WaveSP-Net。
原文摘要 · Abstract (English)
Modern front-end design for speech deepfake detection relies on full fine-tuning of large pre-trained models like XLSR. However, this approach is not parameter-efficient and may lead to suboptimal generalization to realistic, in-the-wild data types. To address these limitations, we introduce a new family of parameter-efficient front-ends that fuse prompt-tuning with classical signal processing transforms. These include FourierPT-XLSR, which uses the Fourier Transform, and two variants based on the Wavelet Transform: WSPT-XLSR and Partial-WSPT-XLSR. We further propose WaveSP-Net, a novel architecture combining a Partial-WSPT-XLSR front-end and a bidirectional Mamba-based back-end. This design injects multi-resolution features into the prompt embeddings, which enhances the localization of subtle synthetic artifacts without altering the frozen XLSR parameters. Experimental results demonstrate that WaveSP-Net outperforms several state-of-the-art models on two new and challenging benchmarks, Deepfake-Eval-2024 and SpoofCeleb, with low trainable parameters and notable performance gains. The code and models are available at https://github.com/xxuan-acoustics/WaveSP-Net.
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。