Deep Convolution Neural Network Model for Facial Expression Analysis Using FastAI

Observing facial expressions aids in perceiving others' emotions, encompassing fear, contempt, disgust, anger, surprise, sadness, happiness, and neutrality. Recognizing subtle emotional variations poses a challenge due to facial expression similarities. It's crucial to acknowledge that emotional displays can vary among individuals experiencing the same feeling. Emotion detection within systems presents a formidable task, but its accuracy holds significance in understanding individuals' emotional states. This technology extends to security applications, adjusting facial expressions in response to different conditions. Errors in the system may manifest as signs of tension in facial expressions. Neural networks and machine learning are effectively employed, particularly the Deep Convolution Neural Network (DCNN) with the LeNet Architecture. The study uses the Kaggle face expression dataset to evaluate the proposed model's performance with metrics like FI Score, Sensitivity, Specificity, and accuracy.