期刊名称：Indian Journal of Computer Science and Engineering
印刷版ISSN：2231-3850
电子版ISSN：0976-5166
出版年度：2021
卷号：12
期号：4
页码：921-944
DOI：10.21817/indjcse/2021/v12i4/211204159
语种：English
出版社：Engg Journals Publications
摘要：While artificial intelligence and machine learning is penetrating through its applications in all the domains, detection of human facial expression has significant role to deal with machines for the interaction. The study is for the development of detection of facial emotion recognition systems are heavily dependent on geometry and appearance. The emotion analysis from images overlooks the challenges of unbalanced dataset and training models to overfitting, it led to a misprediction of actual emotions to be classified. Proposed approach integrates the simplicity of execution, lower computational complexity, exceedingly competitive outcomes beyond numerous real-world emotion classification tasks. The proposed architecture uses ensemble learning techniques on the top of SVM, HMM, KNN, Haar Cascade, Random Forest for emotion classification. Evaluation is done on the popular datasets Ferg-dB, CK+, JAFFE and Fer2013. Study shall predict the universal human emotions more efficiently and precisely. CNN is used for emotion predictor as a model.
关键词：Wireless network;Point-to-multipoint;Internet service
