Optimize Approach to Voice Recognition Using IoT

Verbal communication is one of the best way to behaviors of proclamation for humans, to give and take thought, feeling and relevant data. Speech is the most capable way to teach a device or talk with a device. Detection systems lies on hidden Markov models is doing well less than exacting situation, due to experience as of chief limitations that maximum value of applications of ASR technology in real-world surroundings. However, with in the past decade, several attempt includes to estimate the HMM deficiency. Artificial Neural Networks (ANN) and more specifically Multi-Layer Perception's (MLP) appeared to be a promising alternative in this respect to replace or help HMM in the classification mode. Algorithms are applied to reduce the noise interference and silence suppression. The signal free from above interference is then processed to extract the features. MFCC is used as a feature extraction technique. It will be applied through raspberry pi to device like FAN, BULB through cloud.