To compensate for the inability of the SAF to eliminate noise in diffuse noise fields, we have proposed a hybrid system integrating the SAF system and Wiener filtering (called SAFWF here), and have examined its performance in an isolated nonreverberant sound room [4]. Based on the promising results obtained, we further evaluate the performance of the SAFWF system in real-life mobile and vehicular noisy environments. In this research, we have evaluated the performance of a new hybrid noise reduction system in various real-life noise fields occurring in mobile and vehicular applications. The noise reduction block is composed of a Subband Adaptive Filter (SAF) followed by Wiener filtering. For system evaluation, we have used recordings of real-life noises to contaminate speech materials chosen from the TIMIT database. It is shown that all the recorded noises are samples of diffuse noise fields. As a result, the hybrid system outperforms both the SAF and standard Wiener filtering in all sets of the recordings. The preference of this hybrid system is especially noted in the case of lowpass noise and intense noise conditions. This paper is organized as follows. The structure of the hybrid system is described in the next section. Sec. 3 describes the employed noise and speech materials. System evaluations are reported in Sec. 4, and finally the conclusions of this work are presented in Sec. 5. 2. THE HYBRID SYSTEM
[1]
A.V. Oppenheim,et al.
Enhancement and bandwidth compression of noisy speech
,
1979,
Proceedings of the IEEE.
[2]
J. S. Bird,et al.
Speech enhancement for mobile telephony
,
1990
.
[3]
Hideo Suzuki,et al.
The nature and technology of acoustic space
,
1995
.
[4]
Schuyler Quackenbush,et al.
Objective measures of speech quality
,
1995
.
[5]
R. Brennan,et al.
A flexible filterbank structure for extensive signal manipulations in digital hearing aids
,
1998,
ISCAS '98. Proceedings of the 1998 IEEE International Symposium on Circuits and Systems (Cat. No.98CH36187).
[6]
John H. L. Hansen,et al.
An effective quality evaluation protocol for speech enhancement algorithms
,
1998,
ICSLP.
[7]
Hamid Sheikhzadeh,et al.
Highly oversampled subband adaptive filters for noise cancellation on a low-resource DSP system
,
2002,
INTERSPEECH.
[8]
Hamid Sheikhzadeh,et al.
ETSI AMR-2 VAD: evaluation and ultra low-resource implementation
,
2003,
2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03)..