On the amount of speech data necessary for successful speaker identification

Padrta, Aleš; Radová, Vlasta

Full metadata record

DC Field	Value	Language
dc.contributor.author	Padrta, Aleš
dc.contributor.author	Radová, Vlasta
dc.date.accessioned	2015-12-15T09:41:38Z
dc.date.available	2015-12-15T09:41:38Z
dc.date.issued	2003
dc.identifier.citation	PADRTA, Aleš; RADOVÁ, Vlasta. On the amount of speech data necessary for successful speaker identification. In: Eurospeech 2003 - Interspeech: proceedings of the 8th European Conference on Speech Communication and Technology, 1-4 September 2003, Geneva, Switzerland. [Baixas]: ISCA, 2003, p. 301-304. ISSN 1018-4074.	en
dc.identifier.issn	1018-4074
dc.identifier.uri	http://www.kky.zcu.cz/cs/publications/PadrtaA_2003_Ontheamountof_1
dc.identifier.uri	http://hdl.handle.net/11025/17006
dc.description.abstract	Článek pojednává o závislosti mezi identifikací řečníka a množstvím dat pro testování. V článku jsou prezentovány tři identifikační procedury.	cs
dc.format	4 s.	cs
dc.format.mimetype	application/pdf
dc.language.iso	en	en
dc.publisher	ISCA	en
dc.rights	© Aleš Padrta - Vlasta Radová	cs
dc.subject	identifikace řečníka	cs
dc.subject	uzavřená množina	cs
dc.subject	redukce množství testovacích dat	cs
dc.subject	přístup založený na HMM	cs
dc.title	On the amount of speech data necessary for successful speaker identification	en
dc.title.alternative	Porovnání několika procedur verifikace řečníka založených na GMM	cs
dc.type	článek	cs
dc.type	article	en
dc.rights.access	openAccess	en
dc.type.version	publishedVersion	en
dc.description.abstract-translated	The paper deals with the dependence between the speaker identification performance and the amount of test data. Three speaker identification procedures based on hidden Markov models (HMMs) of phonemes are presented here. One, which is quite commonly used in the speaker recognition systems based on HMMs, uses the likelihood of the whole utterance for speaker identification. The other two that are proposed in this paper are based on the majority voting rule. The experiments were performed for two different situations: either both training and test data were obtained from the same channel, or they were obtained from different channels. All experiments show that the proposed speaker identification procedure based on the majority voting rule for sequences of phonemes allows us to reduce the amount of test data necessary for successful speaker identification.	en
dc.subject.translated	speaker identification	en
dc.subject.translated	closed set	en
dc.subject.translated	reduction of amount of test data	en
dc.subject.translated	HMM based approach	en
dc.type.status	Peer-reviewed	en
Appears in Collections:	Články / Articles (KKY)

Files in This Item:

File	Description	Size	Format
PadrtaA_2003_Ontheamountof_1.pdf	Plný text	197,97 kB	Adobe PDF	View/Open

Show simple item record

Please use this identifier to cite or link to this item: http://hdl.handle.net/11025/17006

search

navigation