REAL-TIME AGE AND GENDER ESTIMATION USING A FINE-TUNED DEEP LEARNING MODEL AND OPENCV

Abstract

In today’s digital era, automatic age and gender classification plays a vital role in various applications, particularly with the growing use of social media platforms. Despite recent advancements in facial recognition algorithms, analyzing real-world photographs continues to pose significant challenges. This study leverages Convolutional Neural Networks (CNNs) in conjunction with the Caffe deep learning framework and OpenCV to evaluate the accuracy of age and gender detection. By applying the Haar Cascade technique for initial face detection, the proposed model demonstrates improved performance in recognizing multiple faces within an image and accurately estimating their age and gender. The model was trained using both positive and negative facial image datasets, and its performance was thoroughly evaluated.