Real-Time Age, Gender, and Emotion Detection Using a Guided Module-Based Convolutional Neural Network for Facial Expression Analysis

This study presents an innovative approach to real-time facial expression analysis using a guided module-based convolutional neural network. The proposed methodology simultaneously detects emotions, age, and gender with high accuracy, achieving 95.1% for seven facial emotions. The research contributes to various fields, including healthcare, security, and human-computer interaction. A custom real-time dataset encompassing diverse age groups was created to enhance the model's efficiency. The study conducted an ablation analysis to optimize the architecture's effectiveness. Quantitative and qualitative results demonstrate superior performance compared to existing methods across multiple datasets. The proposed approach outperforms six state-of-the-art models in accurately detecting emotions based on age and gender in real-time scenarios. This research advances the development of explainable deep-learning models for emotion recognition, addressing challenges posed by specialized datasets and facilitating more sophisticated systems for real-time human interaction analysis.