Certain Investigations for Human Emotion Classification with Sugeno Model of ANFIS
R. Sofia1, D. Sivakumar2
1R. Sofia, Department of Electronics and Instrumentation, Annamalai University, Chidambaram, India.
2D. Sivakumar, Department of Electronics and Instrumentation, Annamalai University, Chidambaram, India.
Manuscript received on 05 February 2019 | Revised Manuscript received on 12 February 2019 | Manuscript Published on 13 February 2019 | PP: 427-432 | Volume-8 Issue- 4S February 2019 | Retrieval Number: DS2901028419/2019©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open-access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Emotion detection has always been a challenging task in our day today life. Identifying emotion of a person will be useful in many areas like in medical field, in interviews, in education, in working environment and so on. Human mind can be read in several ways like, by their standing position, by hand keeping position, but emotion detection by using a human face will be a best choice because of its numerous muscle movements for even a small emotion, and moreover hiding a real feel through face is quite difficult. Emotions are basically classified into Happy, sad, anger, disgust, neutral, fear. The aim of this paper is achieving 100% Human emotion detection using Sugeno model in Adaptive Neuro fuzzy interface system (ANFIS). In this work, initially the human face will be detected. From the detected face the eyes, mouth and eyebrows are extracted and for this feature the various dimensions are measured and the ANFIS is trained with these measurement to identify the emotion of a human. And their performance is justified with various performance measures such as confusion matrix, Regression Plot, Mean Absolute Error, Error Plot, and Error Histogram.
Keywords: Sugeno model, ANFIS, Confusion Matrix, Error Histogram, Mean Absolute Error, Error Plot, Regression Plot.
Scope of the Article: Classification