During my master’s program at Georgia Tech, I contributed to the music performance assessment project using score-informed neural networks. The project investigates how incorporating musical score information can improve the reliability and interpretability of performance evaluation systems. I was involved in developing and analyzing neural architectures that compare vocal pitch contours to corresponding musical scores, helping to assess aspects like intonation and timing accuracy. Our work demonstrated that aligning model inputs with score structure leads to more musically meaningful evaluations, particularly for applications in singing pedagogy and music education.