TCG CREST System Description for the DISPLACE-M Challenge
Nikhil, Md · 2026
Read TCG CREST System Description for the DISPLACE-M Challenge by Nikhil, Md (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-02030.txt.
Description
A TCG CREST team describes its speaker diarization system for a medical conversation challenge in this 2026 report. Diarization answers who spoke when, here in noisy rural healthcare dialogues from the DISPLACE-M challenge. Overlap, dialect shifts, and background noise make the naturalistic recordings hard to split by speaker.
The study compares voice activity detectors and clustering methods inside two frameworks. One is a modular pipeline with speaker embeddings from a SpeechBrain model. The other is a hybrid end-to-end neural system built on a pre-trained speech model. Novel spectral clustering variants compete with classic hierarchical clustering across both frames. The neural system cuts diarization error sharply versus the pipeline baseline in phase one analysis. Their best entry pairs that system with hierarchical clustering and wide median filtering, placing fifth of eleven teams.
Catalog record
| CREATOR | Nikhil, Md |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 19844 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.02030