Micro-expression Analysis for Security and Law Enforcment using Deep Learning

Facial micro-expression recognition has become increasingly important in fields such as psychology, security, and human-computer interaction, due to its potential to reveal genuine emotions. Micro-expressions, lasting only 1/25 to 1/5 seconds, are crucial for applications in public safety, criminal investigations, and clinical medicine. These subtle and brief facial movements reveal hidden emotions, making them valuable for understanding human behavior. This paper presents a novel approach utilizing a hybrid Convolutional Neural Networks-Temporal Convolutional Networks (CNN-TCN) model for micro-expression recognition. The CASME2 dataset is used to train and evaluate the model, ensuring a comprehensive analysis of its performance. This study demonstrates the effectiveness of CNN-TCN in addressing the complexities of micro-expression recognition and contributes to advancing emotion recognition technologies, paving the way for improved human-machine interaction systems. It reviews the advancements in deep learning methods, discusses their challenges, and outlines future trends to guide ongoing research in this field.