. Facial Emotion Recognition — AI Deep Learning Model

I built and deployed a complete deep learning system that can look at a person's face and tell you their emotion — happy, sad, angry, and more — in real time. The project covered the entire pipeline: I built a custom image-preprocessing system from scratch to clean and prepare over 35,000 facial images, then trained and fine-tuned a deep learning model (EfficientNetB0) to classify six different emotions accurately, even with an imbalanced dataset. I then deployed the finished model as a live, public web app where anyone can upload a photo and get an instant result.

This kind of system is useful for any business wanting to understand customer reactions automatically — retail, marketing research, user-experience testing, or any product that benefits from reading facial expressions.

Tools used: Python, TensorFlow/Keras, OpenCV, scikit-learn