An Analysis of the RNN-Based Spoken Term Detection Training
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F49777513%3A23520%2F17%3A43932644" target="_blank" >RIV/49777513:23520/17:43932644 - isvavai.cz</a>
Result on the web
<a href="https://link.springer.com/chapter/10.1007%2F978-3-319-66429-3_11" target="_blank" >https://link.springer.com/chapter/10.1007%2F978-3-319-66429-3_11</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1007/978-3-319-66429-3_11" target="_blank" >10.1007/978-3-319-66429-3_11</a>
Alternative languages
Result language
angličtina
Original language name
An Analysis of the RNN-Based Spoken Term Detection Training
Original language description
This paper studies the training process of the recurrent neural networks used in the spoken term detection (STD) task. The method used in the paper employ two jointly trained Siamese networks using unsupervised data. The grapheme representation of a searched term and the phoneme realization of a putative hit are projected into the pronunciation embedding space using such networks. The score is estimated as relative distance of these embeddings. The paper studies the influence of different loss functions, amount of unsupervised data and the meta-parameters on the performance of the STD system.
Czech name
—
Czech description
—
Classification
Type
D - Article in proceedings
CEP classification
—
OECD FORD branch
20205 - Automation and control systems
Result continuities
Project
<a href="/en/project/TE01020197" target="_blank" >TE01020197: Centre for Applied Cybernetics 3</a><br>
Continuities
P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)
Others
Publication year
2017
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Article name in the collection
Speech and Computer 19th International Conference, SPECOM 2017, Hatfield, UK, September 12-16, 2017, Proceedings
ISBN
978-3-319-66428-6
ISSN
0302-9743
e-ISSN
neuvedeno
Number of pages
11
Pages from-to
119-129
Publisher name
Springer
Place of publication
Cham
Event location
Hatfield, Hertfordshire, United Kingdom
Event date
Sep 12, 2017
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
—