Fusion of diverse denoising systems for robust automatic speech recognition

Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26220%2F14%3APU111005" target="_blank" >RIV/00216305:26220/14:PU111005 - isvavai.cz</a>
Result on the web
<a href="http://ieeexplore.ieee.org/xpl/articleDetails.jsp?tp=&arnumber=6854666&sortType%3Ddesc_p_Publication_Year%26queryText%3DDrotar" target="_blank" >http://ieeexplore.ieee.org/xpl/articleDetails.jsp?tp=&arnumber=6854666&sortType%3Ddesc_p_Publication_Year%26queryText%3DDrotar</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1109/ICASSP.2014.6854666" target="_blank" >10.1109/ICASSP.2014.6854666</a>

Result language
angličtina
Original language name
Fusion of diverse denoising systems for robust automatic speech recognition
Original language description
We present a framework for combining different denoising front-ends for robust speech enhancement for recognition in noisy conditions. This is contrasted against results of optimally fusing diverse parameter settings for a single denoising algorithm. Allfrontends in the latter case exploit the same denoising algorithm, which combines harmonic decomposition, with noise estimation and spectral subtraction. The set of associated parameters involved in these steps are dependent on the noise conditions. Rather than explicitly tuning them, we suggest a strategy that tries to account for the trade-off between average word error rate and diversity to find an optimal subset of these parameter settings. We present the results on Aurora4 database and also compare against traditional speech enhancement methods e.g. Wiener filtering and spectral subtraction.
Czech name
—
Czech description
—

Project
<a href="/en/project/ED2.1.00%2F03.0072" target="_blank" >ED2.1.00/03.0072: Centre of sensor, information and communication systems</a><br>
Continuities
P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Publication year
2014
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection
Acoustics, Speech and Signal Processing (ICASSP), 2014 IEEE International Conference on
ISBN
978-1-4673-5690-9
ISSN
—
e-ISSN
—
Number of pages
4
Pages from-to
5557-5561
Publisher name
IEEE
Place of publication
Florence
Event location
Florence
Event date
May 4, 2014
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
000343655305118

Similar results(10)