Speech is the expression of ideas and emotion through sound. Due to the nature that speech carries emotion, it is a medium for the listeners to identify feelings and attitude of the speaker. It plays a vital role in daily communication among humans. As the demand of human-computer interaction increases, Speech Emotion Recognition (SER) is commonly used to extract emotional state through processed and classified speech signals. However, correctly recognising human emotion from speech is a complex and challenging task since emotions are subjective.
This report aims to detect and identify substantive emotion in voice recording using deep Learning.