Baby Monitoring Through MATLAB Graphical User Interface

This paper describes a novel approach to monitor a baby and it’s emotion and needs. Feature extraction methods like Magnitude Sum function, Pitch and Energy have been performed to classify the signal. These extraction techniques are proven to be more accurate than the conventional techniques. Although combinations of all three techniques have to be used to achieve 100% accuracy, the computational cost and processing time is less than that of Mel Frequency Cepstral Coefficient. Thus, classification of hungry, tired and in-pain cries were successfully done.