International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 96 - Number 13 |
Year of Publication: 2014 |
Authors: Kaladharan N |
10.5120/16858-6739 |
Kaladharan N . Speech Enhancement by Spectral Subtraction Method. International Journal of Computer Applications. 96, 13 ( June 2014), 45-48. DOI=10.5120/16858-6739
Speech enhancement aims to improve speech quality and intelligibility by using various techniques and algorithms. Speech signal is always accompanied with some background noises. Speech processing and communication systems are to apply effective noise reduction techniques in order to extract the desired speech signal from its corrupted speech signal. That is, removal of background noise in the noisy speech. Some of noise reduction techniques are used in the speech processing like spectral subtraction, ceptral mean subtraction, blind equalization ,Adaptive wiener filtering ,Kalman filtering etc. , are used various enhancement situations. Among this spectral subtraction is oldest one of the first algorithm proposed for removal of background noise. It is a single channel speech enhancement method for enhancement of speech degraded locale noise. The locale noise can disturb our conversation in a noisy environment like auditorium, street, market etc. This paper presents the performance of spectral subtraction algorithm is evaluate of a speech by signal to noise ratio value. Spectral subtraction algorithm is widely used in individual conversation due to its simplicity and implementation.