Faster RCNN for Radio Galaxy Detection
Instructions for installation can be found here.
Download and extract RadioGalaxyNET data from here. We expect the directory structure to be the following:
./RadioGalaxyNET/
annotations/ # annotation json files
train/ # train images
val/ # val images
test/ # test images
To train on a single node with single gpu run:
python train.py
To evaluate on test images with a single GPU run:
python evaluate.py
MIT license.