DocumentCode :
3351292
Title :
Sound source separation by using matched beamforming and time-frequency masking
Author :
Beh, Jounghoon ; Lee, Taekjin ; Han, David ; Ko, Hanseok
Author_Institution :
UMIACS, Univ. of Maryland, College Park, MD, USA
fYear :
2010
fDate :
18-22 Oct. 2010
Firstpage :
458
Lastpage :
463
Abstract :
This paper proposes a two-stage algorithm to separate two sound sources by using matched beamforming and time-frequency masking techniques. At first, beamforming was used to separate the sound mixtures back to the original sources while preserving the original contents to the maximum extent. The residual interference was then suppressed by the time-frequency masking technique. A sequential least squares method was used in developing a matched beamformer to estimate the relative transfer function (RTF). From experimental results, it has been shown that the proposed method exhibits improved performance in sound source separation compared to conventional methods. Signal enhanced factor (SEF) was improved by an average of 8.39 dB over the baseline.
Keywords :
array signal processing; human-robot interaction; independent component analysis; least squares approximations; source separation; speech processing; independent component analysis; matched beamforming technique; relative transfer function; residual interference; sequential least squares method; signal enhanced factor; sound source separation; speech based human-robot interface; time-frequency masking technique;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Intelligent Robots and Systems (IROS), 2010 IEEE/RSJ International Conference on
Conference_Location :
Taipei
ISSN :
2153-0858
Print_ISBN :
978-1-4244-6674-0
Type :
conf
DOI :
10.1109/IROS.2010.5652512
Filename :
5652512
Link To Document :
بازگشت