Visual Emotion Recognition Using Deep Neural Networks

被引：0

作者：

Iliev, Alexander I. ^{[1
,2
]}

Mote, Ameya ^{[1
]}

机构：

[1] SRH Univ Berlin, Berlin, Germany

[2] Bulgarian Acad Sci, Inst Math & Informat, Sofia, Bulgaria

来源：

DIGITAL PRESENTATION AND PRESERVATION OF CULTURAL AND SCIENTIFIC HERITAGE | 2022年 / 12卷

关键词：

Emotion Recognition; Image Analysis; Language Understanding; Deep Convolutional Neural Networks; Cross-Cultural Comparison;

D O I：

暂无

中图分类号：

G25 [图书馆学、图书馆事业]; G35 [情报学、情报工作];

学科分类号：

1205 ; 120501 ;

摘要：

It has been proven historically how important feelings and expressions are. They form an important role in communications between individuals of different culture. In the present day, Globalization has led to exchange of vast number of ideas among people on Earth. This gives rise to a unique challenge of identifying what the person in front is speaking about and formulate opinions likewise. Failing to do that would often result in unfortunate consequences. This paper leads to make inroads in this field and provide a basis to other future researchers. We took images from a pre-existing video dataset and recognize the emotions behind it. Through a series of experiments, a final neural network model was created which gave an accuracy of 88%.

引用

页码：77 / 88

页数：12