Skip to content

A web-application to recommend and play music based on facial expressions utilizing a deep learning model.

License

Notifications You must be signed in to change notification settings

codedmachine111/musify

Repository files navigation

Musify

Musify is an innovative web application that combines the power of deep learning and image processing technologies to detect facial emotions and provide personalized music recommendations. Whether you're feeling happy, sad, excited, or calm, Musify delivers a seamless and immersive experience, bringing together the worlds of facial emotion detection and music.
Live Demo | Jupyter Notebook

Features

  • Emotion Detection: Initially utilized a custom-built CNN model for facial emotion detection, and now leverages the highly accurate facial expression model by face-api-js for even better results.
  • Personalized Music Recommendations: Integrates with the Spotify API to curate customized music playlists based on users' facial expression.
  • Face Detection: Employs face-api.js for face detection, ensuring a seamless user experience.

Results of the CNN model

  • Accuracy and loss
  • Confusion matrix

Tools used

                 

  • React: JavaScript library for building user interfaces.
  • Tensorflow: Provides comprehensive ecosystem for building and deploying machine learning/deep learning models.
  • Tensorflow.js: JavaScript library that allows running TensorFlow models directly in the web browser or on Node.js
  • Sass: CSS extension language that provides more advanced features and capabilities.

Installation steps

    • Fork the repo
    • Clone the repo to your local machine git clone https://github.com/codedmachine111/musify.git
    • Change current directory cd musify
  1. Install latest version of Nodejs and install all the dependencies using:
npm install
  1. For using spotify in the web-app, create a .env file in the root directory of the project and add:
VITE_SPOTIFY_CLIENT_ID = "YOUR-SPOTIFY-CLIENT-ID"
VITE_SPOTIFY_CLIENT_SECRET = "YOUR-SPOTIFY-CLIENT-SECRET"
VITE_APP_URL = "VITE-APP-URL-AFTER-HOSTING"

Note You need to get your Spotify ID and SECRET key from here

  1. Run the development server:
npm run dev

Contribution

Contributions are welcome! If you have any suggestions, improvements, or bug fixes, please submit a pull request or open an issue on the GitHub repository.

License

This project is licensed under the MIT License.

About

A web-application to recommend and play music based on facial expressions utilizing a deep learning model.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published