Portfolio Detail

Developed a facial emotion recognition system using Convolutional Neural Network (CNN) with LeNet architecture to analyze and classify human emotions from facial expressions. This project addresses the significance of emotion recognition in various fields such as biomedical engineering, psychology, and workplace management, particularly focusing on mitigating employee stress and turnover.

Utilizing the FER 2013 dataset from Kaggle, which comprises 30,000 grayscale images, the model was trained to identify seven distinct emotional states with an accuracy of 76% and a validation accuracy of 81%. The system demonstrates the applicability of CNNs for emotion classification, offering valuable insights for human-computer interaction and integrating emerging AI and ML technologies to enhance emotional understanding and management.