dc.contributor.author | Van Wyk, Lucas | |
dc.contributor.author | Davel, Marelie, H | |
dc.contributor.author | Van Heerden, Charl | |
dc.date.accessioned | 2022-10-27T19:43:02Z | |
dc.date.available | 2022-10-27T19:43:02Z | |
dc.date.issued | 2021 | |
dc.identifier.citation | Van Wyk , L et al. Unsupervised Fine-tuning of Speaker Diarisation Pipelines using Silhouette Coefficients, volume 11: 202-216.[https://engineering.nwu.ac.za/must-deep-learning/publications] | en_US |
dc.identifier.uri | http://hdl.handle.net/10394/40042 | |
dc.description.abstract | We investigate the use of silhouette coefficients in cluster analysis for speaker diarisation, with the dual purpose of unsupervised fine-tuning during domain adaptation and determining the number of speakers in an audio file. Our main contribution is to demonstrate the use of silhouette coefficients to perform per-file domain adaptation, which we show to deliver an improvement over per-corpus domain adaptation. Secondly, we show that this method of silhouette-based cluster analysis can be used to accurately determine more than one hyperparameter at the same time. Finally, we propose a novel method for calculating the silhouette coefficient of clusters using a PLDA score matrix as input | en_US |
dc.language.iso | en | en_US |
dc.publisher | SACAIR | en_US |
dc.subject | Speaker Diarisation | en_US |
dc.subject | · Unsupervised Fine-Tuning | en_US |
dc.subject | · Domain Adaptation | en_US |
dc.title | Unsupervised Fine-tuning of Speaker Diarisation Pipelines using Silhouette Coefficients | en_US |
dc.type | Article | en_US |