期刊名称：International Journal of Computer Science and Engineering
印刷版ISSN：2278-9960
电子版ISSN：2278-9979
出版年度：2020
卷号：7
期号：6
页码：20-23
DOI：10.14445/23488387/IJCSE-V7I6P106
出版社：IASET Journals
摘要：It is general fact that humans are highlighted by other species in the world by the emotions which is expressed through facial expressions. As the Artificial intelligence are emerging and growing exponentially in many fields where it is important to recognize the human emotions when the auto reply system. The aim of this paper is to propose a system that can be used to recognize the facial expressions of human. An automated emotion recognition system which examines the expressions through various steps like segmentation, feature extraction and identification of human emotions from image or video. This application inspired by image processing and machine learning algorithms. Steps involved in this process are like Image Pre-processing, face detection, facial components, feature extraction and classification. This application uses the image captured from the webcam and the obtained image is compared with the trained dataset model available and then emotional state of the image will be displayed. Facial Emotion Recognition application is implemented using Convolution Neural Network (CNN). System has been tested with the dataset which contains the various emotions of the humans. The application has achieved 56.77 % accuracy and 0.57 precision on testing dataset. This application is capable of classifying human facial expressions into 6 essential emotions: happy, sad, fear, anger, disgust, and neutral.
关键词：Artificial Intelligence; Image Processing; Machine Learning; Face Detection.
