SyncIn-Bring music closer to you

My Final Project

My final year project is SyncIn- Brings music closer to you.
This is a project which is an emotion or a mood-based music player. In this project, my team and I have created an extension that scans or reads the mood of the user or client and accordingly displays the music playlist to the user.
This application not only uses inbuilt face recognition technology of the camera but also diminishes the need for switching in between music apps to get good quality music.
The name of the project SyncIn itself suggests that our moods are incorporated or synced with the music which will be played.
This project tends to be effective in its performance and also tends to be visually interactive and appealing to the user.

Link to Code

mmudit30 / SyncIn

Music Player using Emotion Recognition

How I built it

This project is built by using techniques of face recognition, emotion detection, and music incorporation.
These techniques especially face recognition and emotion detection uses technologies like Open CV and TensorFlow.
The whole project is built on Python.
The project is divided into 3 phases:-

• 1st phase- Emotion detection The initial phase mainly deals with face recognition and emotion detection. In this, we collected a fair amount of data sets of faces by applying to various universities and then trained them so as to recognize a face using Mobilenet v-2. Then we collected some samples of data of every emotion. For now, 3 emotions are evaluated namely Angry, Sad/Neutral, Happy. Each and every emotion's dataset consisted of pictures of each emotion and then were trained. The emotion which was detected was on the basis of the confidence rate. The emotion having the highest confidence rate was the final emotion of the user.
• 2nd phase- Music Incorporation In the second phase, we created datasets of music or songs according to the emotions prescribed. Each of the datasets was created in CSV file with the format of '.mp3'. Now when the face got detected the emotion was mapped to each of the emotion-based music playlists. Once the appropriate playlist got recognized, any random song got played from that playlist. We used the pygame library for python to use the music player. Also, one can pause, resume, stop, and exit the extension.

Additional Thoughts

For the final phase, we plan on enhancing the visual interaction and appeal of the project and make an extension for the browser. For now, we are running the application in a command terminal. We plan to move to a GUI application and add color to every song being played so as to contrast with the user's mood.
This project is a revolutionary phase as it diminishes the distance between music and technology hence enhancing and taking a step forward in the world of automation and human comfort. So, in any case, searching for a music playlist for yourself maybe a thing in the past.