Real-time Facial Recognition and Emotion Detection System

This paper presents a deep learning-based model for real-time automatic mood estimation usingfacial expressions in images. The model is built on a Convolutional Neural Network (CNN) architecture,customized to learn individual facial parameters and map them into facial Action Units (AUs). Theseparameters are then translated into the Pleasure, Arousal, and Dominance (PAD) space, which forms thebasis for mood categorization. The experimental framework defines four primary mood categories:"Exalted", "Calm", "Anxious", and "Bored", based on the Pleasure–Arousal (PA) plane, along withadditional categories for positive and negative Pleasure states. The model's performance is evaluated on astimulus video shown to participants, where their facial expressions are recorded and analyzed. Resultsdemonstrate that the CNN-based model achieves a 94% accuracy in classifying moods in the Pleasuredimension, and 73% accuracy in the PA categorization, highlighting the model's ability to accuratelyestimate moods based on facial expressions. The findings suggest that facial expressions are a reliableindicator of subjective emotional states, offering potential applications in real-time mood assessmentsystems for diverse fields such as human-computer interaction, healthcare, and user experience.