Project for SkelCon
    ├── code:core code for contribution in our paper
        ├── color_space_mixture.py (Data Augmentation Method)  
        ├── sample_contrastive_learning.py (Sample Contrastive Learning)  
        ├── model_skelcon.py (Skeletal Prior based Network)  
        └── ...  
    ├── docs (figures)  
        └── ... 
    ├── onnx (trained weights)  
        ├── *.onnx (Pytorch trained weights)  
        ├── infer.py (to extract vessels from fundus images with *.onnx) 
        └── ...     
    ├── proj (package for segmentation with torch)  
        ├── data (to extract datasets)  
        ├── nets (define the network)  
        ├── build.py (define the network)  
        ├── grad.py (for training)  
        ├── loop.py (for training)  
        ├── optim.py (optimizer)  
        ├── main.py   
        └── ...  
    ├── results (segmentation for fundus images on testsets)  
        ├── popular (segmentation results for popular datasets)  
        ├── generalization (segmentation results for cross-dataset-validation)  
        └── ...