Train an Object Detector using Tensorflow 2 Object Detection API in 2021

I'll describe at a top level what you need to do here, as this part hasn't really changed much from other tutorials. In summary, we are going to download the Egohands dataset but only use a subset of the many many images there, since we are doing transfer learning. We will split them into a train directory and a test directory, and generate .xml First we need to make sure you are in your root directory and then clone my git repo. From my downloaded repo, copy the egohands_dataset_to_csv.py file into your root location and run it.