Garuda - Garba Rujukan Digital

Klasifikasi Ekspresi Wajah Menggunakan Covolutional Neural Network Taufiq Akbar, Ahmad; Akbar, Ahmad Taufiq; Saifullah, Shoffan; Prapcoyo, Hari
Jurnal Teknologi Informasi dan Ilmu Komputer Vol 11 No 6: Desember 2024
Publisher : Fakultas Ilmu Komputer, Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25126/jtiik.2024118888

Pengenalan ekspresi wajah adalah tantangan penting dalam pengolahan citra dan interaksi manusia-komputer karena kompleksitas dan variasi yang ada. Penelitian ini mengusulkan arsitektur sederhana Convolutional Neural Network (CNN) untuk meningkatkan efisiensi klasifikasi emosi pada dataset kecil. Dataset yang digunakan adalah Jaffe, yang terdiri dari 213 citra berukuran 256x256 piksel dalam tujuh kategori ekspresi. Citra-citra tersebut di-resize menjadi 128x128 piksel untuk mempercepat pemrosesan. Data diproses menggunakan arsitektur CNN yang terdiri dari 3 lapisan konvolusi, 2 lapisan subsampling, dan 2 lapisan dense. Kami mengevaluasi model dengan 5-fold dan 10-fold cross-validation untuk estimasi kinerja yang robust, serta teknik hold-out (70:30, 80:20, 85:15, dan 90:10) untuk perbandingan hasil yang jelas. Hasil menunjukkan akurasi tertinggi sebesar 90.6% dengan learning rate 0.001 pada pembagian 85% data latih dan 15% data uji, melebihi model yang lebih kompleks. Meskipun tidak menggunakan transfer learning atau augmentasi data, model ini tetap unggul dibandingkan pendekatan tradisional seperti Local Binary Pattern (LBP) dan Histogram Oriented Gradient (HOG). Dengan demikian, arsitektur CNN yang sederhana ini terbukti efektif untuk pengenalan ekspresi wajah pada dataset kecil. Abstract Facial expression recognition is a significant challenge in image processing and human-computer interaction due to its inherent complexity and variability. This study proposes a simple Convolutional Neural Network (CNN) architecture to enhance the efficiency of emotion classification on small datasets. Jaffe's dataset consists of 213 images sized 256x256 pixels across seven expression categories. These images were resized to 128x128 pixels to accelerate processing. The data was processed using a CNN architecture comprising 3 convolutional layers, 2 subsampling layers, and 2 dense layers. We evaluated the model with 5-fold- and 10-fold cross-validation for robust performance estimation and hold-out techniques (70:30, 80:20, 85:15, and 90:10) for clear result comparison. The results indicated the highest accuracy of 90.6% with a learning rate of 0.001 using the 85% training and 15% testing data split, surpassing that of more complex models. Although the model does not employ transfer learning or data augmentation, it still outperforms traditional approaches such as Local Binary Pattern (LBP) and Histogram Oriented Gradient (HOG). Thus, this simple CNN architecture proves effective for facial expression recognition on small datasets.

Klasifikasi Ekspresi Wajah Menggunakan Covolutional Neural Network Taufiq Akbar, Ahmad; Akbar, Ahmad Taufiq; Saifullah, Shoffan; Prapcoyo, Hari
Jurnal Teknologi Informasi dan Ilmu Komputer Vol 11 No 6: Desember 2024
Publisher : Fakultas Ilmu Komputer, Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25126/jtiik.2024118888

Pengenalan ekspresi wajah adalah tantangan penting dalam pengolahan citra dan interaksi manusia-komputer karena kompleksitas dan variasi yang ada. Penelitian ini mengusulkan arsitektur sederhana Convolutional Neural Network (CNN) untuk meningkatkan efisiensi klasifikasi emosi pada dataset kecil. Dataset yang digunakan adalah Jaffe, yang terdiri dari 213 citra berukuran 256x256 piksel dalam tujuh kategori ekspresi. Citra-citra tersebut di-resize menjadi 128x128 piksel untuk mempercepat pemrosesan. Data diproses menggunakan arsitektur CNN yang terdiri dari 3 lapisan konvolusi, 2 lapisan subsampling, dan 2 lapisan dense. Kami mengevaluasi model dengan 5-fold dan 10-fold cross-validation untuk estimasi kinerja yang robust, serta teknik hold-out (70:30, 80:20, 85:15, dan 90:10) untuk perbandingan hasil yang jelas. Hasil menunjukkan akurasi tertinggi sebesar 90.6% dengan learning rate 0.001 pada pembagian 85% data latih dan 15% data uji, melebihi model yang lebih kompleks. Meskipun tidak menggunakan transfer learning atau augmentasi data, model ini tetap unggul dibandingkan pendekatan tradisional seperti Local Binary Pattern (LBP) dan Histogram Oriented Gradient (HOG). Dengan demikian, arsitektur CNN yang sederhana ini terbukti efektif untuk pengenalan ekspresi wajah pada dataset kecil. Abstract Facial expression recognition is a significant challenge in image processing and human-computer interaction due to its inherent complexity and variability. This study proposes a simple Convolutional Neural Network (CNN) architecture to enhance the efficiency of emotion classification on small datasets. Jaffe's dataset consists of 213 images sized 256x256 pixels across seven expression categories. These images were resized to 128x128 pixels to accelerate processing. The data was processed using a CNN architecture comprising 3 convolutional layers, 2 subsampling layers, and 2 dense layers. We evaluated the model with 5-fold- and 10-fold cross-validation for robust performance estimation and hold-out techniques (70:30, 80:20, 85:15, and 90:10) for clear result comparison. The results indicated the highest accuracy of 90.6% with a learning rate of 0.001 using the 85% training and 15% testing data split, surpassing that of more complex models. Although the model does not employ transfer learning or data augmentation, it still outperforms traditional approaches such as Local Binary Pattern (LBP) and Histogram Oriented Gradient (HOG). Thus, this simple CNN architecture proves effective for facial expression recognition on small datasets.

PELATIHAN PERAKITAN KOMPUTER PADA MA IBNU SINA KUTONTEGALTIRTO BERBAH SLEMAN YOGYAKARTA Budi Suyanto; Ahmad Taufiq Akbar; M. Rafid Zufar
Jurnal Padamu Negeri Vol. 1 No. 4 (2024): Oktober : Jurnal Padamu Negeri (JPN)
Publisher : CV. Denasya Smart Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.69714/44gyrc19

Computers come from the word computer (to compute) which means to calculate. However, in accordance with the development of the times, computers have experienced many improvements, computers do not only function to calculate, but have been used to do various jobs and human activities, such as research, banking, and even education. With the help of computers and other supporting devices such as operating systems and applications, all types of work can be carried out quickly, precisely, and accurately. In completing these tasks, the computer does it automatically based on the instructions given to it. Computer assembly focuses on the introduction and assembly of hardware, to the installation of the operating system. Computer assembly training at MA IBNU SINA Kuton Berbah Sleman Yogyakarta is a community service activity carried out to help students learn more about computer maintenance and use.

EfficientNet B0 Feature Extraction with L2-SVM Classification for Robust Facial Expression Recognition Akbar, Ahmad Taufiq; Saifullah, Shoffan; Prapcoyo, Hari; Rustamadji, Heru; Cahyana, Nur Heri
Journal of Information System and Informatics Vol 7 No 2 (2025): June
Publisher : Universitas Bina Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51519/journalisi.v7i2.1071

Facial expression recognition (FER) remains a challenging task due to the subtle visual variations between emotional categories and the constraints of small, controlled datasets. Traditional deep learning approaches often require extensive training, large-scale datasets, and data augmentation to achieve robust generalization. To overcome these limitations, this paper proposes a hybrid FER framework that combines EfficientNet B0 as a deep feature extractor with an L2-regularized Support Vector Machine (L2-SVM) classifier. The model is designed to operate effectively on limited data without the need for end-to-end fine-tuning or augmentation, offering a lightweight and efficient solution for resource-constrained environments. Experimental results on the JAFFE and CK+ benchmark datasets demonstrate the proposed method’s strong performance, achieving up to 100% accuracy across various hold-out splits (90:10, 80:20, 70:30) and 99.8% accuracy under 5-fold cross-validation. Evaluation metrics including precision, recall, and F1-score consistently exceeded 95% across all emotion classes. Confusion matrix analysis revealed perfect classification of high-intensity emotions such as Happiness and Surprise, while minor misclassifications occurred in more ambiguous expressions like Fear and Sadness. These results validate the model’s generalization ability, efficiency, and suitability for real-time FER tasks. Future work will extend the framework to in-the-wild datasets and incorporate model explainability techniques to improve interpretability in practical deployment Keywords: Facial Expression Recognition, EfficientNet, SVM, Deep Features, Emotion Classification

Title

Found 14 Documents
Search

Abstract

Abstract

Abstract

Abstract

Title Search

Found 14 Documents Search

Abstract

Abstract

Abstract

Abstract

Title

Found 14 Documents
Search