A ML classifier which classifies images based on different emotions (happy, sad, disgust, surprise, neutral). Analyzes frames from video or web cam and predicts an emotion if a face is detected.
The emotion classifier has an accuracy of about 67%, but the performance is quite poor. Currently working on improving this performace.
- Download the repository
- Install dependancies (opencv, tensorflow and numpy)
- Set file path or enable web cam in app.py
- Run python3 app.py