The system implements deep learning model training, image preprocessing, and a real-time inference pipeline using Python. It classifies emotions into Angry, Sad, Fear, Surprise, Neutral, and Disgust from live video feeds. Focused on improving model accuracy and optimizing performance for real-time constraints. Built for training and research purposes.
Real-time webcam-based emotion detection
CNN model trained on facial expression dataset
Classifies 7 emotion categories
Image preprocessing pipeline
Real-time inference optimization
Deep learning model training workflow


