PENGENALAN EMOSI RAUT WAJAH MANUSIA MELALUI REKAMAN VIDEO MENGGUNAKAN METODE CONVOLUTIONAL NEURAL NETWORK (CNN)

JAMES RIO SASUWUK (2023) PENGENALAN EMOSI RAUT WAJAH MANUSIA MELALUI REKAMAN VIDEO MENGGUNAKAN METODE CONVOLUTIONAL NEURAL NETWORK (CNN). Undergraduate Theses thesis, Universitas Tadulako.

Full text not available from this repository.

Abstract

The Covid-19 pandemic has now become a very heartbreaking disaster for all humans on this earth. All activities and lives are certainly disrupted due to this epidemic. The field of education is also one of the obstacles, starting from the kindergarten, elementary, junior high and high school levels, even in the world of lectures, they must carry out the online teaching and learning process. Human emotions are mainly expressed by facial expressions, which are the result of facial muscle movements. Facial expression is a form of nonverbal communication that is expressed through people's minds. This system will classify emotions from facial expressions using the CNN method and detect faces using the Viola-Jones method. In designing this system, we will use the VGG-16 architecture with a layer depth of 16 which consists of a convolutional layer using the ReLU activation function, using a pooling layer, which is more precisely max pooling, then finally using a Fully- Connected Layer using the ReLU activation function and for The output layer will use the softmax activation function. Viola-Jones is a method for detecting facial objects in images and CNN is a method for recognizing facial expressions. The Viola-Jones method has an accuracy of 87.6% for detecting the location of a face and the CNN method has an accuracy of 59.8% for classifying emotions on the face.

Item Type: Thesis (Undergraduate Theses)
Subjects: University Structure > Faculty of Engineering > Informatics Engineering
S1 - Undergraduate Thesis > Faculty of Engineering > Informatics Engineering
Divisions: Faculty of Engineering > Informatics Engineering
Tadulako Subject Areas > S1 - Undergraduate Thesis > Faculty of Engineering > Informatics Engineering
Depositing User: system estd estd
Date Deposited: 01 Feb 2023 07:03
Last Modified: 01 Feb 2023 06:35
URI: http://repository.untad.ac.id/id/eprint/12602

Actions (login required)

View Item View Item