Deteksi ekspresi wajah secara real time menggunakan tensorflow CCN

Salim, Michael (2022) Deteksi ekspresi wajah secara real time menggunakan tensorflow CCN. Bachelor thesis, Universitas Pelita Harapan.

Preview	Text (Title) Title.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (33kB) | Preview
Preview	Text (Abstract) Abstract.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (241kB) | Preview
Preview	Text (ToC) ToC.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (521kB) | Preview
Preview	Text (Chapter1) Chapter1.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (802kB) | Preview
	Text (Chapter2) Chapter2.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (2MB)
	Text (Chapter3) Chapter3.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (1MB)
	Text (Chapter4) Chapter4.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (4MB)
	Text (Chapter5) Chapter5.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (238kB)
Preview	Text (Bibliography) Bibliography.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (212kB) | Preview
	Text (Appendices) Appendices.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (3MB)

Abstract

In social interaction, humans intuitively use facial expressions to establish communication and show their emotions to someone. Facial expression recognition through computers can be used for facial recognition attendance. The identification process of facial expressions can be done conventionally or through applications by applying algorithms. The conventional way of identifying facial expressions is to look at someone's expression and conclude their expression. This conventional process is certainly more accurate but it requires quite a lot of time to identify facial objects. The process of identifying facial expressions through applications is carried out through previous studies. Previous research has a good level of accuracy, but previous research has not tried the application of facial expression identification in real time. Due to the problems and short comings in previous studies, a facial expression detection application will be designed that can be used in real time. This application is built using Tensorflow and implements Convolutional Neural Network (CNN). The test results with Confusion Matrix that the Tensorflow CNN algorithm from the combination of the training data model used is quite accurate in detecting the position of the human face where the test results get an accuracy of 74.17%, precision of 74.07%, and recall of 74.18%./ Dalam berinteraksi sosial, manusia secara intuitif menggunakan ekspresi wajah untuk menjalin komunikasi dan menunjukkan emosinya kepada seseorang. Pengenalan ekspresi wajah melalui komputer dapat digunakan seperti absensi pengenalan wajah atau untuk mengenali kepribadian seseorang. Proses identifikasi ekspresi wajah dapat dilakukan secara konvensional ataupun melalui aplikasi dengan menerapkan algoritma. Cara konvesional dalam identifikasi ekspresi wajah adalah dengan melihat ekspresi seseorang kemudian menyimpulkan ekspresinya. Proses konvensional ini tentunya lebih akurat namun kurang efisien, dikarenakan memerlukan waktu yang cukup lama untuk mengidentifikasi objek wajah. Proses identifikasi ekspresi wajah melalui aplikasi dilakukan melalui penelitian-penelitian terdahulu yang merancang aplikasi deteksi wajah dengan berbagai algoritma. Penelitian terdahulu sudah cukup baik dan memiliki tingkat akurasi yang bagus, namun penelitian terdahulu belum mencoba penerapan identifikasi ekspresi wajah secara real time. Oleh karena permasalahan dan kekurangan pada penelitian sebelumnya, maka akan dirancang sebuah aplikasi pendeteksian ekspresi wajah yang dapat digunakan secara real time. Aplikasi ini dibangun menggunakan Tensorflow dan mengimplementasikan Convolutional Neural Network (CNN). Hasil pengujian dengan Confusion Matrix bahwa algoritma Tensorflow CNN dari kombinasi model data training yang digunakan tergolong cukup akurat dalam mendeteksi posisi wajah manusia dimana hasil pengujian mendapatkan akurasi sebesar 74,17%, presisi sebesar 74,07%, dan recall sebesar 74,18%.

Actions (login required)

View Item