Project for Berkeley Deep RL course: using deep reinforcement learning for segmentation of medical images
Data pre-processing
For the data pre-processing script to work:
Clone cocoapi inside the deeprl_segmentation folder, and follow the instructions to install it (usually just need to run Make inside the PythonAPI folder)
Download your coco dataset (for example, val2017) inside the deeprl_segmentation folder
Download the corresponding annotations, and place them inside a folder called annotations inside the deeprl_segmentation folder