MUS-EMO: Emotion Based Music Recommendation System

This project is aimed to develop the Facial Emotion based Music Recommendation system using Convolution Neural Network.

Project Details

1. To approach the Facial Emotion Analysis, face detection is the priority. The face detection is approached using frontal face haarcascade.
2. After detecting the face, the face image is converted into black & white image and feeded into the CNN model.
3. Convolution Neural Network uses the image of the user and identifies the emotion.
4. The current project version can only detects the 5 emotions - Happy, Sad, Neutral, Excited, Angry.
5. The whole model is then embedded into the Flask Framework.
6. Emotion Analysis accuracy is 73.53%.

Technology Used

Languages: Python, HTML, CSS Web-Framework: Flask

Link to the research paper - Mus-Emo: An Automated Facial Emotion-Based Music Recommendation System Using Convolutional Neural Network