OPTIMIZE APPROACH TO VOICE RECOGNITION

Speech is the most efficient way to train a machine or communicate with a machine. This work focuses on the objective to recognize the word or the phase spoken by human, keywords in high speed. Recognition systems based on hidden Markov models are effective under particular circumstances, but do suffer from some major limitations that limit applicability of ASR technology in real-world environments. However, over the last few years, several attempts have been undergone to evaluate the HMM deficiencies. Artificial Neural Networks (ANN) and more specifically multilayer Perceptron’s (MLP) appeared to be a promising alternative in this respect to replace or help HMM in the classification mode. But ANNs were unsuccessful in dealing with long time sequences of speech signals. So taking the advantages of both the systems into consideration it was proposed to combine HMM and ANN within a single, hybrid architecture. The goal in hybrid systems for ASR is to take advantage from the properties of both HMM and ANNs, improving flexibility and ASR performanc