CMILP

A description projection for our Cross Margin-Information-based Pseudo-Label

Data Preprocess

To prepare the dataset, you can follow the work of WLSMIS

You can also run python code/scribbles_generator.py for scribble label generating and

python code/dataloaders/word_data_processing.py for data preprocessing.

Our work is based on both the 3D and 2D medical volumes.

The splits are available at ./data/WORD.

Model Training

Run

CUDA_VISIBLE_DEVICES=0 python train_ours_refine.py --fold fold1 --seed 114514 --base_lr 3e-2 --num_classes 15 --batch_size 8 --exp WORD/ours_crosspse

for model training. Have fun.