Článek v časopise

FÉR Radek, MATĚJKA Pavel, GRÉZL František, PLCHOT Oldřich, VESELÝ Karel a ČERNOCKÝ Jan. Multilingually Trained Bottleneck Features in Spoken Language Recognition. Computer Speech and Language. Amsterdam: Elsevier Science, 2017, roč. 2017, č. 46, s. 252-267. ISSN 0885-2308. Dostupné z: http://www.sciencedirect.com/science/article/pii/S0885230816302947
Jazyk publikace:angličtina
Název publikace:Multilingually Trained Bottleneck Features in Spoken Language Recognition
Název (cs):Vícejazyčně trénované parametry založené na úzkém hrdle neuronových sítí pro rozpoznávání mluveného jazyka
Strany:252-267
Místo vydání:NL
Rok:2017
URL:http://www.sciencedirect.com/science/article/pii/S0885230816302947
Časopis:Computer Speech and Language, roč. 2017, č. 46, Amsterdam, NL
ISSN:0885-2308
DOI:10.1016/j.csl.2017.06.008
URL:http://www.fit.vutbr.cz/research/groups/speech/publi/2017/fer_CSL2017.pdf [PDF]
Klíčová slova
Multilingual training, Bottleneck features, Spoken language recognition
Anotace
Článek pojednává o vícejazyčně trénovaných parametrech založených na úzkém hrdle neuronových sítí pro rozpoznávání mluveného jazyka.
Abstrakt
Multilingual training of neural networks has proven to be simple yet effective way to deal with multilingual training corpora. It allows to use several resources to jointly train a language independent representation of features, which can be encoded into low-dimensional feature set by embedding narrow bottleneck layer to the network. In this paper, we analyze such features on the task of spoken language recognition (SLR), focusing on practical aspects of training bottleneck networks and analyzing their integration in SLR. By comparing properties of mono and multilingual features we show the suitability of multilingual training for SLR. The state-of-the-art performance of these features is demonstrated on the NIST LRE09 database.
BibTeX:
@ARTICLE{
   author = {Radek F{\'{e}}r and Pavel Mat{\v{e}}jka and
	Franti{\v{s}}ek Gr{\'{e}}zl and Old{\v{r}}ich
	Plchot and Karel Vesel{\'{y}} and Jan
	{\v{C}}ernock{\'{y}}},
   title = {Multilingually Trained Bottleneck Features in
	Spoken Language Recognition},
   pages = {252--267},
   journal = {Computer Speech and Language},
   volume = {2017},
   number = {46},
   year = {2017},
   ISSN = {0885-2308},
   doi = {10.1016/j.csl.2017.06.008},
   language = {english},
   url = {http://www.fit.vutbr.cz/research/view_pub.php.cs?id=11518}
}

Vaše IPv4 adresa: 35.175.248.25