AL_SSDA_WSI

The codes are used for the experiments of MICCAI2022
2022卒-備瀬研-荒木健吾

Directory

• analysis:
  To analyze about the dataset.
• clustering:
  Extract feature vectors from source trained model, and calculate cluster entropy.
• preprocess:
  To crop WSI into patches and split WSIs into train/valid/test.
• S:
  The codes for the experiment S that use only labeled source.
• ST:
  The codes for the experiment S+T that use labeled source and labeled target.
• visualyze:
  some functions for visualyzing featuremap and results of cluster entropy.