You can access the GitHub repository here

You can access the project slides here

You can see our YouTube video here

The h5 file for the model weights and architecture can be accessed here: https://drive.google.com/file/d/1rS6icpUGQwIGMecRsOKuQvZAhKBV2SYu/view?usp=sharing

Project: A personalized, content and action recommendation system based on the detected emotions of the specific user (driver).

Project components include:

• A custom-built emotion recognition model that detects basic emotions from facial expressions in real-time (uses video/webcam input)
• Face ID and recognition
• Dataframe of user preferences and pictures
• Voice agent (uses a text-to-speech Python library)
• Recommendation functions, and using the YouTube API and JavaScript to provide music recommendations
• A final, simple web application made using Flask

1. python 3.6 or above

libraries and packages:

1. face-recognition
2. opencv-python
3. numpy
4. pandas
5. flask
6. tensorflow (2.0.0)
7. pyttsx3
8. imutils