![]() | Only 14 pages are availabe for public view |
Abstract The ultimate goal of speech recognition is to understand the input appropriate response in the classification of words In kindergarten. This thesis is concerned with isolated word recognition systems based on template Matching using the artificial neural network algorithm. End point detection is importance for practical speech recognition to distinguish between noise and speech, Also to reduce the dimension of the Input vector to neural network It was suggested new methods using two techniques called (linear predictive coding & Wavelet analysis) have been studied |