Deep Learning Based Gender Prediction from Eye Images: A Comparative Study Using EfficientNet and MobileNet
DOI:
https://doi.org/10.35718/iiair.v2i1.8481446Keywords:
Gender Prediction, Eye Images, EfficienNet, MobileNet, Deep Learning, Machine Learning, Image ClassificationAbstract
Eye images offer a unique and rich sources of biometric information, with consistent patterns across the eyebrows, eyes, and eyelids, making them ideal for tasks like gender prediction. Gender classification through eye images is an approach for identification without the need for physical contact, making it suitable for device such as on resource constrained device like CCTV or smartphones. This study shows a comparative analysis of two lightweight deep learning architectures, EfficientNet and MobileNet, to identify which one is the optimal and best model. In the evaluation, we used diverse datasets consisting of 13.024 images from Kaggle and additional data from manual documentation. The experimental process in this study consist of several stages, namely data preprocessing, augmentation, model training using adjusted hyperparameters, and evaluation using accuracy, precision, recall, and F1-score. The results showed that EfficientNetB0 had better performance, reaching the highest test accuracy of 96.55%. EfficientNetB0 also showed good balance in classification with F1-scores of 0.97 for males and 0.96 for females, indicating that EfficientNetB0 performes better than MobileNetV3 because it is able to balance computational efficiency with predictive accuracy in classifying gender based on eye image.
References
[1] A. Arifandi. “Identifikasi Dan Prediksi Umur Serta Jenis Kelamin Berdasarkan Citra Wajah Menggunakan Algoritma Convolutional Neural Network (CNN)”,JurnalSains dan Teknologi-Universitas PGRI Kanjuruhan Malang, vol. 4, no. 2, p. 2022, doi: 10.21067/jtst.v4i2.6985.
[2] S. Hosgurmath, V. V. Mallappa, N. B. Patil, and V. Petli, “A face recognition system using convolutional feature extraction with linear collaborative discriminant regression classification,”Int. J. Electr. Comput. Eng., vol. 12, no. 2, p. 1468, 2022, doi: 10.11591/ijece.v12i2.pp1468-1476.
[3] Mozhdeh Rouhsedaghat, Yifan Wang, Xiou Ge, Shuowen Hu, Suya You, and C.-C. Jay Kuo, "FaceHop: A Light-Weight Low-Resolution Face Gender Classification Method" 2020, doi: 10.1007/978-3-030-68793-9_12.
[4] Qingqiao Hua, Siyang Yina, Huiyang Ni, Yisiyuan Huanga, "An End to End Deep Neural Network for Iris Recognition", 2020, doi: 10.1016/j.procs.2020.06.118.
[5] T. Cut Al-Saidina Zulkhaidi, E. Maria, P. “Pengenalan Pola Bentuk Wajah dengan OpenCV,” Studi Teknologi Rekayasa Perangkat Lunak, and P. Pertanian Negeri Samarinda (JURTI), vol. 3, no. 2, 2019, doi: 10.30872/jurti.v3i2.4033.
[6] K. Mujib, A. Hidayatno, and D. T. Prakoso, “Pengenalan Wajah Menggunakan Local Binary Pattern (LBP) Dan Support Vector Machine (SVM)”, 2018.
[7] A. Wulan, “Deep Learning Untuk Deteksi Wajah Yang Berhijab Menggunakan Algoritma Cnn Dengan Tensorflow”, 2020.
[8] M. Naurah,“Pengenalan Wajah Dengan Latar Belakang Bebas Menggunakan Template Matching Dan Local Banry Pattern (LPB)”, 2019.
[9] Anggiratih, Endang, et al. "Klasifikasi Penyakit Tanaman Padi Menggunakan Model Deep Learning Efficientnet B3 dengan Transfer Learning." Jurnal Ilmiah SINUS 19.1(2021): 75-83, doi : 10.30646/sinus.v19i1.526.
[10] https://keras.io/2/api/.
[11] S. W. P. Listio, "Performance of deep learning Inception model and MobileNet model on gender prediction through eye image," Sinkron: Jurnal dan Penelitian Teknik Informatika, vol. 7, no. 4, pp. 11887, Oct. 2022, doi: 10.33395/sinkron.v7i4.11887, doi: 0000-0001-8199-3312.
[12] H. Pu and K. Yi, "A Comparative Analysis of EfficientNet and MobileNet Models’ Performance on Limited Datasets: An Example of American Sign Language Alphabet Detection," Highlights in Science, Engineering and Technology, vol. 94, 2024, CMLAI 2024.
[13] Fernando Alonso-Fernandez, Kevin Hernandez Diaz, Silvia Ramis, Francisco J. Perales, Josef Bigun, "Facial Masks and Soft-Biometrics: Leveraging Face Recognition CNNs for Age and Gender Prediction on Mobile Ocular image", 2021, doi: 10.1049/bme2.12046Digital.
[14] H. T. Nguyen, L. T. T. Pham, D. T. Dang, S. N. Huynh, P. H. Dang, dan Q. T. H. Nguyen, "Age Prediction from Facial Images Using Deep Learning Architecture,"Applied Computer Systems, vol. 29, no. 2, pp. 22–29, Dec. 2024.
[15] R. V. Muddaluru, S. R. Thoguluva, S. Prabha, T. K. Reddy, and S. Palaniswamy, "A Comparative Study of Filters and Deep Learning Models to Predict Diabetic Retinopathy," arXiv preprint arXiv:2309.15216, ver. 3, Jan. 9, 2024, doi:
10.48550/arXiv.2309.15216.
[16] Rahmad Firdaus, Joni Satria, Baidarus, "Klasifikasi Jenis Kelamin Berdasarkan Gambar Mata dengan Menggunakan Algoritma Convolutional Neural Network (CNN), 2022, doi: 10.37859/coscitech.v3i3.4360.
[17] Gilang Syah Amadha K, "Klasifikasi Jenis Kelamin Berdasarkan Citra Mata dengan Menggunakan Algoritma K-Nearest Neighbor, 2023.
[18] Priska Choirina, Rosa Andrie Asmara, "Deteksi Jenis Kelamin Berdasarkan Citra Wajah Jarak Jauh dengan Metode Haar Cascade Classifier", 2016, doi: 10.33795/JIP.V2I4.77.
[19] Rosa Andrie Asmara, Bella Sita Andjani, Ulla Delfana Rosiani, Priska Choirina, "Klasifikasi Jenis Kelamin pada Citra Wajah Menggunakan Metode Naive Bayes", 2018, doi: 10.33795/jip.v4i3.209.
[20] Ulla Delfana Rosiani, Rosa Andrie Asmara, Nadhifatul Laeily, "Penerapan Facial Landmark Point untuk Kasifikasi Jenis Kelamin Berdasarkan Citra Wajah", 2019, doi: 10.33795/jip.v6i1.328.
[21] Khairul Sani, Inung Wijayanto S.T.,M.T., Eko usatio S.T.,M.T. "Implementasi Aplikasi Pengenalan Jenis Kelamin Berdasarkan Citra Wajah dengan Metode Support Vector Machine secara Real Time, 2016.
[22] B. A. Kumar and M. Bansal, "Face mask detection on photo and real-time video images using Caffe-MobileNetV2 transfer learning," Applied Sciences, vol. 13, no. 2, p. 935, 2023, doi: 10.3390/app13020935.
[23] M. R. D. Ulhaq, D. Firdaus, and M. A. Zaidan, "Pengenalan ekspresi wajah secara real-time menggunakan metode SSD MobileNet berbasis Android," Journal of Technology and Informatics (JoTI), vol. 5, no. 1, Oct. 2023, doi:
10.48550/arXiv.1905.11946.
[24] M. Tan and Q. V. Le, “EfficientNet: Rethinking model scaling for convolutional neural networks,” arXiv preprint arXiv:1905.11946, version 5, Sep. 2020, doi:
10.48550/arXiv.1905.11946.
[25] Zhou, A., Ma, Y., Ji, W. et al. Multi-head attention-based two-stream EfficientNet for action recognition. Multimedia Systems 29, 487–498 (2023), doi: 10.1007/s00530-022-00961-3.
[26] T. Shahriar, "Comparative Analysis of Lightweight Deep Learning Models for Memory-Constrained Devices," arXiv preprint arXiv:2505.03303, 2025, doi:
10.48550/arXiv.2505.03303.
[27] Karim, Md & Goni, Md & Nahiduzzaman, Md & Ahsan, Mominul & Haider, Julfikar & Kowalski, Marcin, “Enhancing agriculture through real-time grape leaf disease classification via an edge device with a lightweight CNN architecture and Grad-CAM”, 2024, doi: 10.1038/s41598-024-66989-9.
Downloads
Published
How to Cite
License
Copyright (c) 2026 Innovative Informatics and Artificial Intelligence Research
This work is licensed under a Creative Commons Attribution 4.0 International License.
