Neural Network Based Voiced and Unvoiced Classification Using Egg and Mfcc Feature

2 Assistant Professor PET Engineering College, Vallioor ---------------------------------------------------------------------***--------------------------------------------------------------------Abstract Speech recognition is a subjective phenomenon. This procedure still faces a considerable measure of issue. Different techniques are utilized for various purposes. In this work of project, it is shown that how the speech signals is perceived utilizing Fuzzy c-implies (FCM) Clustering Method in neural system. Voices of various people of different ages in a quiet and noise free condition by a good quality receiver are recorded. Same sentence of term 10-12 seconds is talked by these people. These talked sentences are then changed over into wave positions. At that point components of the recorded examples are removed via preparing these signals utilizing LPC. These systems are prepared to perform by the pattern recognition. Their significance to the order and portrayal of composed content is low; be that as it may, most viable speech recognition frameworks depend intensely on speech recognition to accomplish elite. For the vowel classification utilizing adaptive median filter with combination of speech and EGG information. The Mel-Frequency Cepstral Coefficients (MFCC) and Neural Network (NN) are utilized as components representing to the speech signal.