Real-Time Detection Of Human Speech Emotion Using ATMEGA

Abstract

Emotion detection is a process of automatically recognizing the emotion of a speaker based on features. A suitable feature extraction technique can make it possible to detect and inform an individual about his/her emotions.  It can be used for safety and security by alarming an individual of an angry mob or by informing him/ her about the voice tempo during a public speech or presentation. The goal of this work is to build simple yet complete and representative automatic speech emotion recognition system. Due to a limited availability of voice sensors, this work tests the emotional voice samples using the amplitude, energy and pitch features. However, one can have a large amount of database and different sensors to develop a for more accurate system.