用吸引子聚类实现端到端说话人分离，精度高且模型小。
End-to-End Diarization utilizing Attractor Deep Clustering
- 通过变换器更新吸引子，结合交叉注意力和卷积模块优化说话人表示。
- 在4个数据集上达到0.9%~1.7%的低错误率，参数量显著减少。
- 适合需要轻量级、高鲁棒性的实时语音分离场景。
说话人分离仍面临结构化说话人表示、高效建模及对变化条件鲁棒性差的挑战。我们提出一种高性能、紧凑的端到端分离框架，集成Conformer解码器、变换器更新的吸引子及深度聚类风格的角度损失。方法通过增强的Conformer结构改进说话人表示，引入吸引子交叉注意力与额外卷积模块。为强化结构化嵌入，扩展深度聚类以构建标签-吸引子向量，使其方向结构与音频嵌入对齐。同时对活跃吸引子施加正交约束，提升说话人分离效果，并抑制非活跃吸引子以避免误激活。最后采用排列不变的二值交叉熵损失优化说话人检测。实验表明，该方法在保持低参数量的同时实现了低分离错误率。
原文摘要 · Abstract (English)
Speaker diarization remains challenging due to the need for structured speaker representations, efficient modeling, and robustness to varying conditions. We propose a performant, compact diarization framework that integrates conformer decoders, transformer-updated attractors, and a deep clustering style angle loss. Our approach refines speaker representations with an enhanced conformer structure, incorporating cross-attention to attractors and an additional convolution module. To enforce structured embeddings, we extend deep clustering by constructing label-attractor vectors, aligning their directional structure with audio embeddings. We also impose orthogonality constraints on active attractors for better speaker separation while suppressing non-active attractors to prevent false activations. Finally, a permutation invariant training binary cross-entropy loss refines speaker detection. Experiments show that our method achieves low diarization error while maintaining parameter count.
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。