Digital Signal Processing Reference
In-Depth Information
Chapter
11
SPEECH ENHANCEMENT BASED ON F-NORM
CONSTRAINED TRUNCATED SVD
ALGORITHM
Guo Chen, Soo Ngee Koh and Ing Yann Soon
School of Electrical & Electronic Engineering, Nanyang Technological University, Singapore
639798
Email:
esnkoh@ntu.edu.sg
Abstract:
Traditional singular value decomposition (SVD) based speech enhancement
algorithms are usually limited by the use of a fixed order of retained singular
values which may not be optimal for time-varying noise corrupted speech
signals. In this chapter, we propose the use of a Frobenius-norm (F-norm)
constrained truncated (FCTSVD) algorithm in an analysis-by-synthesis
procedure for choosing the appropriate order of retained singular values for
speech enhancement. It allows for self-adaptation in time and for different
noise and noisy speech characteristics. Also, it leads to the best approximation
of original speech in terms of SNR. The proposed algorithm has been tested
and compared with a traditional SVD algorithm for different noise types and
levels. Simulation results show that it achieves higher SNR improvements for
both additive white noise and colored noise as compared to a traditional SVD
algorithm.
Keywords:
Speech enhancement, singular value decomposition (SVD), Frobenius-norm.
1.
INTRODUCTION
The use of speech processing systems for voice communication and
speech recognition is becoming more and more common. This is largely due
to the availability of low cost digital signal processors and memory chips.
Among all the speech processing research efforts, the problem of enhancing
speech degraded by additive broad-band noise is still an active research
