Inspiration

Browsing through songs on daily-basis to only find the one good song suited to us , is indeed a tiring process. Keeping that in mind and using latest technologies, the browsing time can be reduced and thus comes the solution as MoodMuse.

What it does

MoodMuse plays songs on the basis of your mood, the mood being analysed using both visual and audio

How we built it

It is built using python mainly.

Challenges we ran into

The real challenge was setting up the Machine Learning system. Since our systems didn't have GPU support, we decided to run the ML inference on the cloud (Google Collab). The communication between cloud server and the computer server was an issue and it does require a good knowledge of Cloud Computing.

Accomplishments that we're proud of

We learnt about Spotipy (Spotify based API), Mediapipe API.

What we learned

We basically learned about Machine Learning based APIs for the very first time. Also, we learnt how Cloud Computing can be leveraged for the usecase of Machine Learning based applications.

What's next for MoodMuse

Make it Multilingual. Improvement over the algorithm, improve the accuracy of the ML algorithm. Expand to other entertainment domain.

Built With

Share this project:

Updates