Článek ve sborníku konference

ZEINALI Hossein, SAMETI Hossein, BURGET Lukáš, ČERNOCKÝ Jan, MAGHSOODI Nooshin a MATĚJKA Pavel. i-vector/HMM Based Text-dependent Speaker Verification System for RedDots Challenge. In: Proceedings of Interspeech 2016. San Francisco: International Speech Communication Association, 2016, s. 440-444. ISBN 978-1-5108-3313-5. Dostupné z: https://www.researchgate.net/publication/303895014_i-VectorHMM_Based_Text-Dependent_Speaker_Verification_System_for_RedDots_Challenge
Jazyk publikace:angličtina
Název publikace:i-vector/HMM Based Text-dependent Speaker Verification System for RedDots Challenge
Název (cs):Systém pro ověřování mluvčího závislý na textu založený na kombinaci i-vektorů a HMM pro RedDots Challenge
Strany:440-444
Sborník:Proceedings of Interspeech 2016
Konference:Interspeech 2016
Místo vydání:San Francisco, US
Rok:2016
URL:https://www.researchgate.net/publication/303895014_i-VectorHMM_Based_Text-Dependent_Speaker_Verification_System_for_RedDots_Challenge
ISBN:978-1-5108-3313-5
DOI:10.21437/Interspeech.2016-1174
Vydavatel:International Speech Communication Association
URL:http://www.fit.vutbr.cz/research/groups/speech/publi/2016/zeinali_interspeech2016_IS161174.pdf [PDF]
Klíčová slova
text-dependent speaker verification, i-vector, HMM, RedDots challenge
Anotace
Článek pojednává o systému pro ověřování mluvčího závislého na textu, který je založený na kombinaci i-vektorů a HMM pro RedDots Challenge.
Abstrakt
Recently, a new data collection was initiated within the RedDots project in order to evaluate text-dependent and text-prompted speaker recognition technology on data from a wider speaker population and with more realistic noise, channel and phonetic variability. This paper analyses our systems built for RedDots challenge - the effort to collect and compare the initial results on this new evaluation data set obtained at different sites. We use our recently introduced HMM based i-vector approach, where, instead of the traditional GMM, a set of phone specific HMMs is used to collect the sufficient statistics for i-vector extraction. Our systems are trained in a completely phraseindependent way on the data from RSR2015 and Libri speech databases. We compare systems making use of standard cepstral features and their combination with neural network based bottle-neck features. The best results are obtained with a scorelevel fusion of such systems.
BibTeX:
@INPROCEEDINGS{
   author = {Hossein Zeinali and Hossein Sameti and
	Luk{\'{a}}{\v{s}} Burget and Jan
	{\v{C}}ernock{\'{y}} and Nooshin Maghsoodi and
	Pavel Mat{\v{e}}jka},
   title = {i-vector/HMM Based Text-dependent Speaker
	Verification System for RedDots Challenge},
   pages = {440--444},
   booktitle = {Proceedings of Interspeech 2016},
   year = 2016,
   location = {San Francisco, US},
   publisher = {International Speech Communication Association},
   ISBN = {978-1-5108-3313-5},
   doi = {10.21437/Interspeech.2016-1174},
   language = {english},
   url = {http://www.fit.vutbr.cz/research/view_pub.php.cs?id=11268}
}

Vaše IPv4 adresa: 3.81.28.94
Přepnout na https