Developing Radif-AI, a Machine Learning Model for Gushe (Mode) Classification and Generation in Persian Music Repertoire (Radif Dastgahi)

As someone who is deeply passionate about the Persian music repertoire, I’ve spent a couple of years trying to learn and immerse myself in this rich cultural heritage. My journey has led to some rather amusing and humbling experiences, particularly as I tried my hand at playing some of the repertoire’s musical instruments. In hindsight, it’s fair to say that my musical talents have been a source of comic relief for friends and family alike. But despite these minor setbacks, my love for this repertoire remains undiminished.

Now, I find myself at a fascinating crossroads, where I can combine my longstanding passion for Radif repertoire with my expertise in machine learning. Following proposal aims to contribute to the preservation and advancement of this incredible art form by training an ML model on the Radif repertoire. I am excited about the potential applications of the results, from fostering deeper understanding and appreciation of Persian music to promoting cross-cultural dialogue and exchange.

Abstract: In this study, we propose a machine learning framework to classify and generate music samples for Persian Radif Dastgahi, specifically focusing on the Tar instrument performances by Nur Ali Boroumand and Ali Akbar Shahnazi, on behalf of Mirza Abdollah Farahani and Aqa Hossein Qoli Farahani, two well-known co-founders of contemporary Radif among the…