This SuperUROP aims to develop an assessment system for piano performance integrating expert qualitative evaluations with quantitative computational analysis. Performance data will be collected using Disklavier MIDI output and synchronized audio/video recordings, allowing the capture of both technical (e.g., timing and movement) and expressive aspects of performance.

The project will involve collecting performance data, applying computer vision methods, and using Python for MIDI data processing and event detection. An evaluation rubric will be designed using criteria from prestigious competitions and established pedagogical frameworks. By comparing expert assessments with computationally derived metrics, we seek to reveal correlations between human judgment and computational models.

Through this SuperUROP, I want to build hands-on experience in research that connects computer science and the arts. As a pianist, I’m excited to apply my technical, research, and music skills to this project while deepening my understanding of interdisciplinary work. I look forward to learning, collaborating, and working with other pianists in this unique research setting.