Age and gender classification using Seg-Net based architecture and machine learning

A facial recognition framework is a natural face-recognizing process from a computerized image or videos. Nowadays, for real-time applications, i.e., human–computer interaction, visual supervision, commercial applications, etc., Human Facial features are utilized for gender classification (GC) and age classification. This paper focuses on gender and age classification methodology from various face images—the proposed work based on Seg-Net-based architecture with machine learning algorithm gives excellent results. The overall accuracy increased through the advanced Seg-Net architecture and Support Vector Machine for age and gender recognition. Our proposed method achieved better results in Age classification on various datasets, i.e., Adience, IOG, and FG-Net datasets, accuracy 74.5%, 75.7%, and 92.48%, and in GC also achieved better results as compared to existing technology on the various datasets, i.e., Adience, IOG, FEI, and own datasets respectively accuracy 88.3%, 95.1%, 94.1%, and 91.8%.