Feature And Score Level Combination Of Subspace Gaussians In LVCSR Task
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F13%3APU106379" target="_blank" >RIV/00216305:26230/13:PU106379 - isvavai.cz</a>
Result on the web
<a href="http://www.fit.vutbr.cz/research/groups/speech/publi/2013/motlicek_icassp2013_0007604.pdf" target="_blank" >http://www.fit.vutbr.cz/research/groups/speech/publi/2013/motlicek_icassp2013_0007604.pdf</a>
DOI - Digital Object Identifier
—
Alternative languages
Result language
angličtina
Original language name
Feature And Score Level Combination Of Subspace Gaussians In LVCSR Task
Original language description
We have demonstrated that the SGMM framework is an efficient approach in the LVCSR task. Overall evaluations of SGMMs exploiting powerful but complex PLP-BN features yield similar results as those obtained by conventional HMM/GMMs. Nevertheless, the total number of SGMM parameters is about 3 times less than in the HMM/GMM framework. Evaluation results also indicate different properties of the examined acoustic modeling techniques. Although SGMMs consistently outperform HMM/GMMs when built over individual features, HMM/GMMs can benefit much more from the feature-level combination than SGMMs. Nevertheless based on an analysis measuring complementarity of individual recognition systems, we show that SGMM-based recognizers produce heterogeneous outputs (scores) and thus subsequent score-level combination can bring additional improvement.
Czech name
—
Czech description
—
Classification
Type
D - Article in proceedings
CEP classification
—
OECD FORD branch
10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)
Result continuities
Project
<a href="/en/project/GPP202%2F12%2FP604" target="_blank" >GPP202/12/P604: Speech recognition for low-resource languages</a><br>
Continuities
P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)
Others
Publication year
2013
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Article name in the collection
Proceedings of ICASSP 2013
ISBN
978-1-4799-0355-9
ISSN
—
e-ISSN
—
Number of pages
5
Pages from-to
7604-7608
Publisher name
IEEE Signal Processing Society
Place of publication
Vancouver
Event location
Vancouver
Event date
May 27, 2013
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
—