Hi everyone,
Iām really happy to report that our team has successfully completed CP#18, a music genre classifier project, after five weeks of collaborative effort. We developed and deployed several models (with accuracies ranging from 70-90%), including custom architectures and automated selections, using Streamlit and React for interactive user interfaces. This project highlighted the importance of teamwork, iterative development, and clear communication in achieving our goals.
While we're proud of our results, we recognize limitations in handling genre variations, subgenres, and the evolution of musical styles over time. To address this in the future, we're considering:
Synthetic Data: Leveraging synthetic music to capture a wider range of genre characteristics.
Advanced Classification: Exploring hierarchical or multi-label models for improved handling of subgenres and hybrid music styles.
Try Our Apps!
We'd love for you to test out our team members' applications, check out the code, and let us know what you think!
Andy Chenās App: https://music-classifier.andy-chen.dev/
Julien Hovanās App: https://jewelshovan-sds-cp018-music-class-web-appjulien-hovanapp-pxnmij.streamlit.app/
Salma Wahwah's App: https://salmamusic.streamlit.app/
Hemant Gulati's App: https://hemant-sds-classifymusicgenre.streamlit.app/
Rajan Hansā App: https://rajan-music-genre-xgboost.streamlit.app/
A Huge Thanks to the entire team
Andy Chen , Rajat Thakur, Hemant Gulati, Salma Wahwah, Rajan Hans, Anıl Ćakmak, Vaibhav Chopra , Joshua Rosell
I want to extend my sincere gratitude to each team member for their active participation and dedication to improvement over the past five weeks. The collaborative spirit and engaging discussions during our meetings were truly invaluable to the project's success.
A final big shoutout to Shaheer for their awesome guidance and support throughout this project!
Source Code:

Julien Hovan Your guidance throughout this project has been truly inspiring, and Iām eager to continue learning from you in future endeavors. Thank you for being such an exceptional mentor.
Huge congratulations to the whole team Julien Hovan, Andy Chen, Rajat Thakur, Hemant Gulati, Salma Wahwah, Rajan Hans, Anıl Ćakmak, Vaibhav Chopra, Joshua Rosell, and Shaheer Airaj Ahmed! š¤ Not an easy project so well done. All the different apps are super cool! šµ
Hadelin de Ponteves Please try my app as wellš
https://cnn-music-genre-classifier.streamlit.app/