DEEP LEARNING-BASED AGE AND GENDER DETECTION USING CNN AND HYBRID MACHINE LEARNING APPROACHES

Abstract

Age and gender detection using facial images is a critical task in computer vision with applications in personalized marketing, healthcare, surveillance, and human-computer interaction. This paper presents a deep learning-based system utilizing Convolutional Neural Networks (CNN) combined with hybrid machine learning models for accurate age and gender classification from facial images. The proposed framework employs OpenCV-based face detection with HAAR-cascades, followed by feature extraction and classification using CNN and Support Vector Machines (SVM). The UTKFace and FG-NET datasets were used for training and evaluation, ensuring robust testing across diverse age groups and gender categories. Image pre-processing techniques, including mean filtering and alignment, were applied to improve system accuracy and noise reduction. Experimental results demonstrate the model’s effectiveness across different age brackets (children, teenagers, adults, and elderly), achieving consistent gender and age group detection with high accuracy while retaining computational efficiency. The study highlights the potential of combining deep learning with traditional feature extractors and hybrid optimization strategies for scalable, real-time, and ethically responsible age and gender detection systems.