Paper Abstract and Keywords |
Presentation |
2009-07-17 14:30
Blind Signal Extraction Based on Minimum Mean-Square Error Short-Time Spectral Amplitude Estimator Ryoi Okamoto, Yu Takahashi, Hiroshi Saruwatari, Kiyohiro Shikano (Nara Inst. of Sci and Tech.) EA2009-39 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
In this paper, we propose a new blind speech extraction method consisting of a minimum mean-square error short-time spectral amplitude (MMSE STSA) estimator and noise estimation based on independent component analysis. First, we handle a computer simulation using the artificial noise only in which the stationarity is controlled parametrically, and the result indicates that the proposed method is superior to the conventional methods, blind spatial subtraction array and the original MMSE STSA estimator, under the non-point-source and nonstationary noise condition. Finally, we conduct a experiment in an actual railway-station environment, and objective and subjective evaluations are shown to confirm the advantage of the proposed method in real world. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
Blind source separation / Microphone-array signal processing / Independent component analysis / Spectral subtraction / MMSE STSA estimator / / / |
Reference Info. |
IEICE Tech. Rep., vol. 109, no. 136, EA2009-39, pp. 13-18, July 2009. |
Paper # |
EA2009-39 |
Date of Issue |
2009-07-10 (EA) |
ISSN |
Print edition: ISSN 0913-5685 Online edition: ISSN 2432-6380 |
Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
Download PDF |
EA2009-39 |
|