Voice Signals Feature Extraction and Classification of Bedridden Patients

This manuscript explores voice signal feature extraction, classification and its application on bedridden patients. In this manuscript, three deep learning model Convolutional Neural Network (CNN), Recurrent Neural Network Hidden Markov Model (RNN-HMM) and Recurrent Neural Network- Bi- Long Short-Term Memory (RNN-Bi-LSTM) are used for voice signal classification, and Mel – Frequency Cepstral Coefficients (MFCC) and Zero-crossing rate (ZCR) are used to extract features. Voice recognition systems are a vital aspect of human machine interaction (HMI). It is one of the most fascinating ways to perform any physical task verbally, solely through human commands. There are so many methods that have been invented for voice recognition systems but still it is a challenging task among researchers and academicians to find out the best method for voice recognition for bedridden patients. This work is helpful for bedridden patients to make their life easier