A Study on Emotion Detection

Authors

  • Dhammjyoti Dhawase, Shraddha Belhekar , Priya Patil, Snehal Hulule, Tanvi Ghare, Pankaj Shinde

Keywords:

Emotion detection, Facial expression recognition, visual features.

Abstract

Emotion Detection and recognition is the very hot ongoing topic. As many researches work in this field still the work not reaches that much high level. At some extent we are minimizing gap of emotion recognition. We are working on visual features to recognize emotions using Facial Expression Recognition (FER). In this paper, we have performed a literature survey of those models that were proposed earlier by the former scientists and researchers in the past researches. Those models have been trained on different datasets including AffectNet, CK+, FER2013, and so on. We have performed a survey and a detailed analysis of the previously done research based on the Facial Expression Recognition technology. It has ultimately led to a good understanding of the problems and the issues that were faced by the earlier experiments and analysis as well as much new innovative architecture including those CNN, ensembles, and many more.

It has been discovered in this review that only few researches have been working on the eighth emotion category called ‘contempt’. In our study of previous research papers, we found that AffectNet dataset was the largest dataset to handle as it has the emotions that are divided into eight categories and hence it was hardly implemented in the models previously. We are finding that which model can be utilized and can give better accuracy for a particular dataset. ¬Additionally, the FER research must take into account how differently children, adults, and senior citizens express their emotions through their facial expressions. This review paper has application in different domains like analysis of human computer interaction field, medical fertility, educational field and many more. Our main motto is at some extent this survey may be useful for the upcoming innovators.

 

Published

2023-02-22

How to Cite

Dhammjyoti Dhawase, Shraddha Belhekar , Priya Patil, Snehal Hulule, Tanvi Ghare, Pankaj Shinde. (2023). A Study on Emotion Detection. SJIS-P, 35(1), 86–93. Retrieved from http://sjis.scandinavian-iris.org/index.php/sjis/article/view/239

Issue

Section

Articles