Disclaimer This is an initial attempt to upload a local Jupyter Notebook project that I used for this project to GitHub. Currently I doubt that this will run properly online because it relies on some command line tools to be installed and included on the path, but I believe that if the Markdown and code from the notebook are presented the process will become more clear.
The purpose of this project is to use computer vision to analyze a standard 1v1 Super Smash Bros Melee match hosted on YouTube. The user will submit a YT videoId OCR to map each player's % damage over time. This data will be used to interpret the flow of the match graphically, exposing the effectiveness of each player's punish game.