Thai Voice Recognition for Controlling Electrical appliances Using Long Short-Term Memory

Human speech possesses characteristics in each of the word that can be recognized and learned by computers. In this research, It is being proposed the use of the Deep Learning Model to predict speech turn-on and turn-off various electrical appliances, by using the sound conversion method that has been through the process to get the value of sound waves and applied toward training process in different ways. As the sound has more than 1 syllable and having characteristics of similar words that might difficult to predict. This research is based on Convolutional Neural Network (CNN) for comparison with the use of Long Short-Term Memory (LSTM), which is part of the Recurrent Neural Network (RNN) and Thai language Speech Dataset turn-on and turn-off by the 7 types of electrical appliances, the process of reducing noise and silence of the front and back of the audio files by 14 classes in total. The experimental results signify that the proposed Long Short-Term Memory can achieve the best accuracy.