ON SPEECH RECOGNITION PERFORMANCE UNDER NON-STATIONARY ECHO CANCELLATION

Mahdi Triki

2011

Abstract

During the last decades, performance of speech recognizers significantly increased for large vocabulary tasks and adverse environments. To reduce interference, acoustic echo cancellation has been proposed and extensively investigated. Particular attention was paid to the convergence proprieties and the capability to handle double talk. However, in time-varying environment, the echo canceller has the additional task to track the variations of the propagation channel. With this respect, it has been established that algorithms that exhibit fast convergence do not provide necessarily good tracking performances. In such an environment, performance assessment is also challenging and the ‘experiment’ design is crucial to provide consistent and interpretable results. In the present paper, we reproduce time-varying artifacts by altering the surrounding acoustic environment (using a moving person/robot). The movement characteristics (discrete/continuous) and location (line-of-sight/background) emphasizes different room/algorithms characteristics and provides deeper insights on the system behavior.

Paper Citation

in Harvard Style

Triki M. (2011). ON SPEECH RECOGNITION PERFORMANCE UNDER NON-STATIONARY ECHO CANCELLATION . In Proceedings of the International Conference on Bio-inspired Systems and Signal Processing - Volume 1: BIOSIGNALS, (BIOSTEC 2011) ISBN 978-989-8425-35-5, pages 316-321. DOI: 10.5220/0003182903160321

in Bibtex Style

@conference{biosignals11,
author={Mahdi Triki},
title={ON SPEECH RECOGNITION PERFORMANCE UNDER NON-STATIONARY ECHO CANCELLATION},
booktitle={Proceedings of the International Conference on Bio-inspired Systems and Signal Processing - Volume 1: BIOSIGNALS, (BIOSTEC 2011)},
year={2011},
pages={316-321},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0003182903160321},
isbn={978-989-8425-35-5},
}

in EndNote Style

TY - CONF
JO - Proceedings of the International Conference on Bio-inspired Systems and Signal Processing - Volume 1: BIOSIGNALS, (BIOSTEC 2011)
TI - ON SPEECH RECOGNITION PERFORMANCE UNDER NON-STATIONARY ECHO CANCELLATION
SN - 978-989-8425-35-5
AU - Triki M.
PY - 2011
SP - 316
EP - 321
DO - 10.5220/0003182903160321