A CNN-BASED APPROACH FOR DEVELOPING AN AUTOMATED FACE DETECTION SYSTEM

Sari

Facial detection technology has become increasingly significant across various domains, including security, surveillance, biometrics, and human–computer interaction. This study aims to develop a facial detection system using a Convolutional Neural Network (CNN) approach to recognize and classify facial features from image data. A quantitative experimental method was employed, integrating deep learning modeling with software engineering principles. The system was implemented in MATLAB using a graphical user interface (GUI), and the model was trained on a dataset of 800 facial expression images. Data preprocessing involved image resizing, pixel normalization, and data augmentation to enhance model robustness. The dataset was divided into training, validation, and testing sets to ensure reliable model evaluation and generalization. The CNN model was trained and empirically tested to assess its performance under various conditions. The results indicate that the proposed system is capable of accurately detecting facial regions, even under natural lighting and complex backgrounds. The model achieved high accuracy in face identification, as evidenced by correct bounding box localization in the output interface. Furthermore, the system demonstrated stable performance across different facial orientations. However, several limitations remain, including the need for improved multi-face detection, real-time processing optimization, and broader evaluation across diverse expressions and demographic variations. In conclusion, the CNN-based facial detection system demonstrates strong effectiveness and accuracy, indicating its potential applicability in real-world visual recognition systems.

Keywords:  CNN, Face Detection, Deep Learning, Influence. Image Recognitiion, Matlab