For our Spring 2021 semester ITEC 4700 Artificial Intelligence project, the team was tasked with selecting a topic of interest that could apply Machine Learning to generate a solution. After much deliberation, we decided to base our project on a project conducted by four students at Stanford University called Music Genre Classification. http://cs229.stanford.edu/proj2016/report/BurlinCremeLenain-MusicGenreClassification-report.pdf
Our original plan was to follow the concept of classifying .mp3 files into four different categories of music (i.e. Country, Classical, Metal, and Pop) similar to that of our peers at Standford University. However, we quickly realized that we had very little understanding of MFCC coefficients and how they pertained to the algorithms we were utilizing in class. Our team spent a few weeks trying to understand how to incorporate the concept into our relative plan, but due to time constraints we started to discuss alternative approaches to classifying music, or a form of it. Thus, we came up with the idea of using what we knew about image classification to classify the different musical notes that can be found on sheet music.