We followed the documentation on YOLO's website to create a model using custom data but moved to Google Colab for better processing power. Due to time constraints, we used YOLO's Python notebook tutorial for the proper environment for training our custom data. We did not exactly follow the Python notebook but instead uploaded a revised version of it to this repo. It is quite messy and untidy due to time constraints.
Our data set will not be published due to the nature of crowd-sourcing and the privacy of those who volunteered.
Here are some success photos
We will be continuing this project once time allows :)