Metraipolis › archive › dc_7ca356c4053e

Sequence-to-Sequence Refinement for EEND-VC Diarization with a Shared Encoder — arXiv:2611.02709

Source: arXiv (en)
Original: https://arxiv.org/abs/2611.02709
License (as archived): CC0-1.0
Archived: 2026-11-03T11:55:20.844Z
SHA-256 of the archived text: ef3d7befb5fb40254c071beee1d75f2b31ca019b03b4fa4f7fbf960ab75b7a81
Stable URL of this page: https://metraipolis.com/documento/dc_7ca356c4053e

How to cite

Sequence-to-Sequence Refinement for EEND-VC Diarization with a Shared Encoder — arXiv:2611.02709 — arXiv (en), https://arxiv.org/abs/2611.02709 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7ca356c4053e (Metraipolis archive dc_7ca356c4053e, sha256 ef3d7befb5fb40254c071beee1d75f2b31ca019b03b4fa4f7fbf960ab75b7a81)

BibTeX

@misc{dc7ca356c4053e,
  title = {Sequence-to-Sequence Refinement for EEND-VC Diarization with a Shared Encoder — arXiv:2611.02709},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2611.02709},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7ca356c4053e; sha256 ef3d7befb5fb40254c071beee1d75f2b31ca019b03b4fa4f7fbf960ab75b7a81},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.02709. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

End-to-end neural diarization with vector clustering (EEND-VC) separates local segmentation from global speaker assignment, leaving speaker confusion (SC) despite strong segmentation. We propose shared-encoder sequence-to-sequence (S2S) refinement, which reuses DiariZen's self-supervised learning (SSL) representations for speaker-conditioned decoding and prior-constrained assignment. We study frozen-feature refinement with WavLM and structurally pruned w2v-BERT 2.0, and extend the latter to joint training of local segmentation and refinement heads. The heads learn separate layer mixtures while their losses update one acoustic encoder. Across eight domains, frozen-feature refinement reduces macro diarization error rate (DER) from 12.95% to 12.40% with WavLM and from 12.35% to 12.08% with w2v-BERT 2.0, with relative SC reductions of 17.4% and 14.6%. It adds approximately 15% to system real-time factor. The joint system further lowers macro DER to 11.90%, improving seven of eight domains over non-joint refinement. Error decomposition separates SC-driven refinement gains from detection-driven joint-system gains.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv eess.AS, nuovi lavori del giorno).
Categorie arXiv: Audio and Speech Processing (eess.AS).
Pagina del lavoro: https://arxiv.org/abs/2611.02709 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-03T11:15:52.709Z dalla pagina https://arxiv.org/list/eess.AS/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).