Built and trained a CNN from scratch to recognize facial expression. Classifies face based on emotion into Angry, Disgust, Fear, Happy, Sad, Surprise, Neutral. A web interface is used to perform real-time on image and video data. The project can be braodly divided into two parts -
- Build and train a model in Keras for Facial Expression Recognition.
- Deploy the model on web using FLASK and run it on videos.