EEND-SAA: Enrollment-Less Main Speaker Voice Activity Detection Using Self-Attention Attractors
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_ | 1866909788782198784 |
|---|---|
| author | Wu, Wen-Yung Hsieh, Pei-Chin Chi, Tai-Shih |
| author_facet | Wu, Wen-Yung Hsieh, Pei-Chin Chi, Tai-Shih |
| contents | Voice activity detection (VAD) is essential in speech-based systems, but traditional methods detect only speech presence without identifying speakers. Target-speaker VAD (TS-VAD) extends this by detecting the speech of a known speaker using a short enrollment utterance, but this assumption fails in open-domain scenarios such as meetings or customer service calls, where the main speaker is unknown. We propose EEND-SAA, an enrollment-less, streaming-compatible framework for main-speaker VAD, which identifies the primary speaker without prior knowledge. Unlike TS-VAD, our method determines the main speaker as the one who talks more steadily and clearly, based on speech continuity and volume. We build our model on EEND using two self-attention attractors in a Transformer and apply causal masking for real-time use. Experiments on multi-speaker LibriSpeech mixtures show that EEND-SAA reduces main-speaker DER from 6.63% to 3.61% and improves F1 from 0.9667 to 0.9818 over the SA-EEND baseline, achieving state-of-the-art performance under conditions involving speaker overlap and noise. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_11957 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | EEND-SAA: Enrollment-Less Main Speaker Voice Activity Detection Using Self-Attention Attractors Wu, Wen-Yung Hsieh, Pei-Chin Chi, Tai-Shih Audio and Speech Processing Voice activity detection (VAD) is essential in speech-based systems, but traditional methods detect only speech presence without identifying speakers. Target-speaker VAD (TS-VAD) extends this by detecting the speech of a known speaker using a short enrollment utterance, but this assumption fails in open-domain scenarios such as meetings or customer service calls, where the main speaker is unknown. We propose EEND-SAA, an enrollment-less, streaming-compatible framework for main-speaker VAD, which identifies the primary speaker without prior knowledge. Unlike TS-VAD, our method determines the main speaker as the one who talks more steadily and clearly, based on speech continuity and volume. We build our model on EEND using two self-attention attractors in a Transformer and apply causal masking for real-time use. Experiments on multi-speaker LibriSpeech mixtures show that EEND-SAA reduces main-speaker DER from 6.63% to 3.61% and improves F1 from 0.9667 to 0.9818 over the SA-EEND baseline, achieving state-of-the-art performance under conditions involving speaker overlap and noise. |
| title | EEND-SAA: Enrollment-Less Main Speaker Voice Activity Detection Using Self-Attention Attractors |
| topic | Audio and Speech Processing |
| url | https://arxiv.org/abs/2509.11957 |