期刊名称：Current Journal of Applied Science and Technology
印刷版ISSN：2457-1024
出版年度：2021
卷号：40
期号：20
页码：7-16
DOI：10.9734/cjast/2021/v40i2031459
语种：English
出版社：Sciencedomain International
摘要：Facial expressions are a vital part of human life. Each day has a number of instances and all instances include numerous amounts of communication. Every communication expressed with emotion tells us about the state of the person. The interpersonal as well as security purposes are solved through facial expressions. The mischievous intention of a person can be caught by his expressions.The human mind can capture visual information faster. So, a machine recognizing it will be a challenge. As the saying goes- “A picture is worth a thousand words”- only when it is represented well. A machine being able to detect the atmosphere by the means of expression is less of a manual work.This paper detects the faces, extract the features as well classify them into different categories which ultimately lead to expression recognition. We evaluate our proposed method with the dataset which we used and the recall of angry, fear, happy, neutral, sad, and surprise is 60%, 31%, 84%, 22%, 57% and 58% respectively and the f1-score is 51% 35%, 82%, 25%, 51% and 64% respectively. Experimental results demonstrate the competitive classification of our proposed system.
