Word Extraction from Speech Recognition using Correlation Coefficients

Speech is the fundamental way of communicating with one another. It simply refers to transmission of messages. In case of speech production the information is transmitted in the form of analog waveform that can be transmitted, recorded or decoded. A number of algorithms for speech recognition have been proposed. In this paper, we have suggested an innovative approach of speech recognition. We have initially stored some voice in the database where the same speaker has told different words. Then we have inputted a sample voice of the same person through the microphone where he is speaking a specific word which is already stored in the database. We have performed the task of similar word recognition by finding the gray scale image and histogram plot of inputted word and finally we have used the correlation coefficient for making comparison between two words. General Terms Speech Processing, Word Extraction, Correlation Coefficients.