A novel Arabic Speech Recognition method using neural networks and Gaussian Filtering.

Phoneme recognition is a very popular area of research, especially in the field of machine intelligence, as it adds a distinguish touch in technology world in many fields that provides many features, This paper discussed a phoneme recognition using neural network technique by using a strong algorithm in pre-processing stage to get good results. Moreover, it explained the stages of phoneme recognition, pre-processing procedures such as getting a signal, sampling, quantization, determine an energy then using neural network to get a good results and enhancements. Gaussian Low Pass filtering was used to get enhanced results of voice signal and lowering the noisy; and then use neural network in a training stage to train a system to recognize the speech signal, this paper has benefited from some of the previous results for other researchers and entered good ideas to enhance Arabic voice recognition. [ABSTRACT FROM AUTHOR]