We used the Yolo computer vision model to create the MVP because it is open source.

This wouldn't be so accurate if it weren't for this