Redefining the Bayesian information criterion for speaker diarisation

Stafylakis, T; Katsouros, V; Carayannis, G

dc.contributor.author	Stafylakis, T	en
dc.contributor.author	Katsouros, V	en
dc.contributor.author	Carayannis, G	en
dc.date.accessioned	2014-03-01T02:52:06Z
dc.date.available	2014-03-01T02:52:06Z
dc.date.issued	2009	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/35843
dc.relation.uri	http://www.scopus.com/inward/record.url?eid=2-s2.0-70450175271&partnerID=40&md5=070b01691969ae52cdabd65260818957	en
dc.subject	Bayesian information criterion	en
dc.subject	Cluster analysis	en
dc.subject	Speaker diarisation	en
dc.subject.other	Bayesian information criterion	en
dc.subject.other	Benchmark tests	en
dc.subject.other	Dissimilarity measures	en
dc.subject.other	Hierarchical clustering algorithms	en
dc.subject.other	Penalty term	en
dc.subject.other	Sample sizes	en
dc.subject.other	Bayesian networks	en
dc.subject.other	Cluster analysis	en
dc.subject.other	Speech communication	en
dc.subject.other	Clustering algorithms	en
dc.title	Redefining the Bayesian information criterion for speaker diarisation	en
heal.type	conferenceItem	en
heal.publicationDate	2009	en
heal.abstract	A novel approach to the Bayesian Information Criterion (BIC) is introduced. The new criterion redefines the penalty terms of the BIC, such that each parameter is penalized with the effective sample size is trained with. Contrary to Local-BIC, the proposed criterion scores overall clustering hypotheses and therefore is not restricted to hierarchical clustering algorithms. Contrary to Global-BIC, it provides a local dissimilarity measure that depends only the statistics of the examined clusters and not on the overall sample size. We tested our criterion with two benchmark tests and found significant improvement in performance in the speaker diarisation task. Copyright © 2009 ISCA.	en
heal.journalName	Proceedings of the 10th Annual Conference of the International Speech Communication Association, INTERSPEECH 2009	en
dc.identifier.spage	1051	en
dc.identifier.epage	1054	en