Identifikasi Jenis Kelamin Otomatis Berdasarkan Mata Manusia Menggunakan Convolutional Neural Network (CNN) dan Haar Cascade Classifier

Afu Ichsan Pradana (Universitas Duta Bangsa Surakarta, Indonesia)
Wijiyanto Wijiyanto (Universitas Duta Bangsa Surakarta, Indonesia)

G-Tech : Jurnal Teknologi Terapan
Vol 8 No 1 (2024): G-Tech, Vol. 8 No. 1 Januari 2024

In forensics and security, it is necessary to determine a person's gender. Gender identification utilizing several types of identification, such as pictures of faces, voices, or handwriting, has been extensively studied in recent years. But a lot of offenders are hard to spot on surveillance tape because they cover their heads or have masks on that only show particular eye shapes. In this article, we explore the usage of a CNN with Relu activation for each hidden layer and the Haar Cascade Classifier Algorithm to detect objects of the human eye to recognize the human eye using deep learning. 11.525 Images of male and female eyes were used as the study's data. Utilizing Adam's optimization (Adaptive Moment Estimation), the training procedure lasts for 20 epochs. This study's findings have a 92% accuracy rate for automatically identifying gender. The performance evaluation matrix was used in this investigation, and it produced an overall F1-Score of 93%.