search

ucuapps / WSMIS

Weakly Supervised Medical Images Segmentation
GNU General Public License v3.0
30 stars 10 forks source link

Question about IRN implementation #2

Open sgadgil6 opened 3 years ago

sgadgil6 commented 3 years ago

While loading the dataset for training IRNet (PneumothoraxAffinityDataset) in /irn/voc2/dataloader.py, you load the mask on line 324. Is this the same segmentation mask saved on line 80 of cam.py in /inference/cam_generation/cam.py? If so, the original IRNet code had a step for converting cams to ir labels (irn/step/cam_to_ir_label.py). Do you make use of that step for the pneumothorax dataset or do you directly use the segmentation mask for training IRNet?

Mugen-chen commented 3 months ago

I have the same problem. Have you solved it？