DocumentCode :
1707338
Title :
A reproducing kernel Hilbert space approach for speech enhancement
Author :
Gauci, Oliver ; Debono, Carl J. ; Micallef, Paul
Author_Institution :
Dept. of Commun. & Comput. Eng., Univ. of Malta, Msida
fYear :
2008
Firstpage :
831
Lastpage :
835
Abstract :
The problem of speech enhancement has drawn a considerable amount of research attention over the past few years. Among the techniques developed one finds subspace methods, which seem to offer a good compromise between signal distortion and residual noise level. In this contribution, we present a novel subspace approach to single-channel speech enhancement. The eigen decomposition which was originally performed in the input space is now being done in a reproducing kernel Hilbert space, where the speech nonlinearities can be considered. The proposed algorithm was tested in various noise conditions including white, car, pink and train station noises at various signal-to-noise ratios (SNRs). Objective results show that for white noise, the algorithm presents an average improvement of 73.26% while for colored noise an average improvement of 68.42% is achieved. Subjective tests made on speech, corrupted with white and colored noises, demonstrate that the proposed algorithm provides a significant improvement over other speech enhancement methods found in literature.
Keywords :
Hilbert spaces; covariance matrices; eigenvalues and eigenfunctions; principal component analysis; speech enhancement; covariance matrix estimation; kernel PCA; nonlinear eigen decomposition technique; reproducing Kernel Hilbert space approach; single-channel speech enhancement; speech nonlinearities; subspace approach; 1f noise; Colored noise; Distortion; Hilbert space; Kernel; Noise level; Signal to noise ratio; Speech enhancement; Testing; White noise; Karhunen-Louve transform; reproducing kernel Hilbert space; speech enhancement; subspace-based methods;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Communications, Control and Signal Processing, 2008. ISCCSP 2008. 3rd International Symposium on
Conference_Location :
St Julians
Print_ISBN :
978-1-4244-1687-5
Electronic_ISBN :
978-1-4244-1688-2
Type :
conf
DOI :
10.1109/ISCCSP.2008.4537338
Filename :
4537338
Link To Document :
بازگشت