Comparison of similarity among sub-categories of angry speech emotion

Irritation and disgust are two sub-categories of angry emotion that often confuse the detection system based on anger analysis. This work aims to distinguish these secondary emotions from the primary angry state. Recognition and segregation of these states can help the society to maintain a better social relationship. Further, it can be used as an input to the psychologists and medical practitioners and can provide advanced warning before onset of possible full-blown emotion. The traumatic person can be treated with compassion before any unwarranted rupture in social relationship. As these secondary emotions are often similar and overlapped with angry state, a line drawn in between these states will create a favorable platform for effective negotiation. We have attempted to take up this issue based on different similarity measures between these states. Features such as short time Fourier transform (STFT), chirp transform and short time energy (STE) are used to put a boundary among these states. Effective results have been manifested and shown in the result section.