This HTML5 document contains 44 embedded RDF statements represented using HTML+Microdata notation.

The embedded RDF content will be recognized by any processor of HTML5 Microdata.

Namespace Prefixes

PrefixIRI
n10http://linked.opendata.cz/ontology/domain/vavai/riv/typAkce/
dctermshttp://purl.org/dc/terms/
n20http://localhost/temp/predkladatel/
n16http://purl.org/net/nknouf/ns/bibtex#
n15http://linked.opendata.cz/resource/domain/vavai/projekt/
n17http://linked.opendata.cz/resource/domain/vavai/vysledek/RIV%2F46747885%3A24220%2F12%3A%230002004%21RIV13-TA0-24220___/
n13http://linked.opendata.cz/resource/domain/vavai/riv/tvurce/
n12http://linked.opendata.cz/resource/domain/vavai/subjekt/
n9http://linked.opendata.cz/ontology/domain/vavai/
n22https://schema.org/
shttp://schema.org/
skoshttp://www.w3.org/2004/02/skos/core#
n3http://linked.opendata.cz/ontology/domain/vavai/riv/
n2http://linked.opendata.cz/resource/domain/vavai/vysledek/
rdfhttp://www.w3.org/1999/02/22-rdf-syntax-ns#
n8http://linked.opendata.cz/ontology/domain/vavai/riv/klicoveSlovo/
n21http://linked.opendata.cz/ontology/domain/vavai/riv/duvernostUdaju/
xsdhhttp://www.w3.org/2001/XMLSchema#
n11http://linked.opendata.cz/ontology/domain/vavai/riv/jazykVysledku/
n6http://linked.opendata.cz/ontology/domain/vavai/riv/aktivita/
n19http://linked.opendata.cz/ontology/domain/vavai/riv/obor/
n14http://linked.opendata.cz/ontology/domain/vavai/riv/druhVysledku/
n4http://reference.data.gov.uk/id/gregorian-year/

Statements

Subject Item
n2:RIV%2F46747885%3A24220%2F12%3A%230002004%21RIV13-TA0-24220___
rdf:type
skos:Concept n9:Vysledek
dcterms:description
In this paper we present our system for speaker diarization of broadcast news based on recent advances in the speaker recognition field. In the system, speaker segments determined by the speaker changepoint detector are represented by i-vectors and similarity of segments’ speakers evaluated using cosine distance scoring. Linear discriminant analysis is employed to cope with intra-speaker variability. The experiments were carried out using the COST278 multilingual broadcast news database. We demonstrate improvement of the performance over the baseline system based on the Bayesian Information Criterion (BIC) and highlight significant impact of cepstral mean normalization. Finally, two-stage clustering employing BIC-based clustering to pre-cluster segments in the first stage is examined and showed to yield further performance improvement. The best performing configuration of our system achieved 52.4 % relative improvement of the speaker error rate over the baseline. In this paper we present our system for speaker diarization of broadcast news based on recent advances in the speaker recognition field. In the system, speaker segments determined by the speaker changepoint detector are represented by i-vectors and similarity of segments’ speakers evaluated using cosine distance scoring. Linear discriminant analysis is employed to cope with intra-speaker variability. The experiments were carried out using the COST278 multilingual broadcast news database. We demonstrate improvement of the performance over the baseline system based on the Bayesian Information Criterion (BIC) and highlight significant impact of cepstral mean normalization. Finally, two-stage clustering employing BIC-based clustering to pre-cluster segments in the first stage is examined and showed to yield further performance improvement. The best performing configuration of our system achieved 52.4 % relative improvement of the speaker error rate over the baseline.
dcterms:title
Speaker Diarization of Broadcast Streams using Two-stage Clustering based on I-vectors and Cosine Distance Scoring Speaker Diarization of Broadcast Streams using Two-stage Clustering based on I-vectors and Cosine Distance Scoring
skos:prefLabel
Speaker Diarization of Broadcast Streams using Two-stage Clustering based on I-vectors and Cosine Distance Scoring Speaker Diarization of Broadcast Streams using Two-stage Clustering based on I-vectors and Cosine Distance Scoring
skos:notation
RIV/46747885:24220/12:#0002004!RIV13-TA0-24220___
n9:predkladatel
n12:orjk%3A24220
n3:aktivita
n6:P
n3:aktivity
P(TA01011204)
n3:dodaniDat
n4:2013
n3:domaciTvurceVysledku
n13:1170368 n13:8048150
n3:druhVysledku
n14:D
n3:duvernostUdaju
n21:S
n3:entitaPredkladatele
n17:predkladatel
n3:idSjednocenehoVysledku
170198
n3:idVysledku
RIV/46747885:24220/12:#0002004
n3:jazykVysledku
n11:eng
n3:klicovaSlova
speaker diarization; broadcast news; clustering; i-vectors
n3:klicoveSlovo
n8:speaker%20diarization n8:broadcast%20news n8:clustering n8:i-vectors
n3:kontrolniKodProRIV
[BECE5495B368]
n3:mistoKonaniAkce
Tokyo, Japonsko
n3:mistoVydani
Japonsko
n3:nazevZdroje
Proc. of International Conference on Acoustics, Speech, and Signal Processing - ICASSP 2012
n3:obor
n19:JC
n3:pocetDomacichTvurcuVysledku
2
n3:pocetTvurcuVysledku
2
n3:projekt
n15:TA01011204
n3:rokUplatneniVysledku
n4:2012
n3:tvurceVysledku
Pražák, Jan Silovský, Jan
n3:typAkce
n10:WRD
n3:wos
000312381404066
n3:zahajeniAkce
2012-01-01+01:00
s:numberOfPages
4
n16:hasPublisher
Neuveden
n22:isbn
978-1-4673-0046-9
n20:organizacniJednotka
24220