End-to-End Diarization utilizing Attractor Deep Clustering
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912427288821760 |
|---|---|
| author | Palzer, David Maciejewski, Matthew Fosler-Lussier, Eric |
| author_facet | Palzer, David Maciejewski, Matthew Fosler-Lussier, Eric |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_11090 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | End-to-End Diarization utilizing Attractor Deep Clustering Palzer, David Maciejewski, Matthew Fosler-Lussier, Eric Sound Audio and Speech Processing 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. |
| title | End-to-End Diarization utilizing Attractor Deep Clustering |
| topic | Sound Audio and Speech Processing |
| url | https://arxiv.org/abs/2506.11090 |