SCorP MIUA2024

SCorP: Statistics-Informed Dense Correspondence Prediction Directly from Unsegmented Medical Images

To run training and inference

python run_train.py --config configs/liver.json

To run with point cloud data, replace idx to None in the run_train.py

Data organization:

train
	- meshes 
	- images 
val
	- meshes
	- images 
test 
	- meshes
	- images 
mean.particles  

TODO: Upload dataset to SW Cloud portal and provide links