Estimasi Dimensi Kepribadian Melalui Analisis Citra Ekspresi Wajah Menggunakan Convolutional Neural Network

Abstract

Personality is a crucial aspect that influences a person's behavior, way of thinking, and interaction patterns in various situations. Advances in digital image processing and artificial intelligence technology offer opportunities for developing systems capable of automatically estimating personality dimensions through facial expression analysis. This study aims to build a personality dimension estimation model based on facial expression images using the Convolutional Neural Network (CNN) method. The research steps include collecting a dataset of facial images representing various expressions, image preprocessing, including face detection, image size normalization, and data augmentation, followed by training a CNN model to learn visual characteristics related to facial expressions. The resulting model is then tested using data not involved in the training process to measure the model's generalization ability. System performance is evaluated using metrics such as accuracy, precision, recall, F1-score, and confusion matrix. The results are expected to demonstrate that the Convolutional Neural Network approach is capable of effectively extracting visual features from facial expressions and can therefore be used as a basis for estimating personality dimensions. This research is expected to contribute to the development of computer vision and artificial intelligence technology, particularly in the fields of human behavior analysis, decision support systems, and more adaptive human-computer interaction applications.

Keywords: Personality Estimation, Facial Expression Images, Digital Image Processing, Convolutional Neural Network, Computer Vision, Artificial Intelligence.