This thesis investigates an implementation of speech recognition front-end. It is an application specific integrated circuit (ASIC) solution. A Mel Cepstrum algorithm is implemented for the feature extraction. We prese ...