编解码器转换中存留的信息：用于语音深度伪造检测的池化无伴奏残差
What Survives the Codec Shift: Pooled No-Vocals Residuals for Speech Deepfake Detection
浏览论文内容
中文总结 AI 辅助
本研究针对编解码器转换下语音深度伪造检测泛化难的问题，提出双视角检测器MN-P，融合池化无伴奏残差与XLS-R特征，使EER总体降低54.2%。
中文摘要 AI 辅助
从基于声码器的语音合成向神经编解码器语音合成的转变，使得语音深度伪造检测器的泛化更加困难，尤其是那些依赖面向语音表示的检测器。当生成机制发生变化时，哪些声学表示仍保留判别性信息尚不清楚。因此，我们使用一个共享的低容量线性分类器比较了12种声学表示，以识别在编解码器转换下有效的证据。分析表明，层次化XLS-R在汇总测试集上表现最佳，而池化无伴奏残差统计在未见编解码器条件下表现最优，揭示了不同生成条件下的互补行为。基于这一发现，我们提出了MN-P，一种双视角检测器，通过自适应门控将话语级池化无伴奏表示与词元级XLS-R特征相结合。所提出的MN-P相对于表现最佳的重新训练的最先进系统，在总体上将等错误率（EER）降低了54.2%，在编解码器未见条件下降低了60.9%，并在不同检测器后端上均取得了一致的提升。这些结果表明，池化无伴奏残差统计为跨生成语音深度伪造检测提供了有效的互补证据。
英文摘要
The transition from vocoder-based to neural-codec speech synthesis makes generalization more difficult for speech deepfake detectors, particularly those relying on speech-oriented representations. It remains unclear which acoustic representations retain discriminative information when the generation mechanism changes. We therefore compare 12 acoustic representations using a shared low-capacity linear classifier to identify effective evidence under codec shift. The analysis shows that hierarchical XLS-R leads on the pooled test set, while pooled no-vocals residual statistics perform best on the unseen-codec condition, revealing complementary behavior across generation conditions. Building on this finding, we propose MN-P, a dual-view detector that integrates an utterance-level pooled no-vocals representation with token-level XLS-R features through adaptive gating. The proposed MN-P reduces EER by 54.2% overall and by 60.9% on the codec-unseen condition relative to the best-performing retrained state-of-the-art system, with consistent gains across different detector backends. These results indicate that pooled no-vocals residual statistics provide effective complementary evidence for cross-generation speech deepfake detection.
发表机构
- Tsinghua University（清华大学）
机构由 AI 辅助整理，请以论文原文为准。
