SUDx: Speaker-conditioned U-shaped Diarization with Speaker Extraction-guided Enhancement

(TASLP 2026)

Update Log

14/11/2025 – Added the SUDx model architecture. Additional components and documentation will be progressively refined and expanded.

Name		Name	Last commit message	Last commit date
Latest commit History 2 Commits
configs		configs
src		src
LICENSE.txt		LICENSE.txt
README.md		README.md
requirements.txt		requirements.txt