Applying Machine Learning Techniques for Speech Emotion Recognition

Emotion is an instinctive or intuitive feeling which is distinguished from reasoning or knowledge, it is a strong feeling derived from one's circumstance or surroundings. The main idea of the paper is to apply Deep Neural Network (DNN) and k-nearest neighbor (k-NN) in recognition of emotion from speech-especially scary state of mind. The area of application of the system is mainly concerned over the health care units. The foundation of this research has its main firm applications in palliative care. Under most precise outcome the alert signals are made through cloud. Many raw data are collected under special emphasis techniques. The acoustic voice signals are converted to wave form, uttereance level feature extraction emotion classification, existing database recognition, alert signal creation through cloud is the sequence of steps to be followed. The findings of the paper lays a fruitful contribution to palliative care system.