How to train the SegAN?

[NeuZhangQiang]

Thank you very much for sharing SegAN. Currently, I apply the SegAN to segment the rib. If I use the Unet with the resnet as the backbone, the dice can achieve 0.85. But for SegAN, I can only obtain dice of 0.7. Do you have any suggestion?

[blublebee]

Traceback (most recent call last):
File "D:\PycharmFile\ImageSegmentation\GAN\SegAN-master\train.py", line 123, in
for i, data in tqdm(enumerate(dataloader, 1)):
File "E:\anaconda\envs\pytorch\lib\site-packages\tqdm\std.py", line 1195, in iter
for obj in iterable:
File "E:\anaconda\envs\pytorch\lib\site-packages\torch\utils\data\dataloader.py", line 628, in next
data = self._next_data()
File "E:\anaconda\envs\pytorch\lib\site-packages\torch\utils\data\dataloader.py", line 671, in _next_data
data = self._dataset_fetcher.fetch(index) # may raise StopIteration
File "E:\anaconda\envs\pytorch\lib\site-packages\torch\utils\data_utils\fetch.py", line 58, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File "E:\anaconda\envs\pytorch\lib\site-packages\torch\utils\data_utils\fetch.py", line 58, in
data = [self.dataset[idx] for idx in possibly_batched_index]
File "D:\PycharmFile\ImageSegmentation\GAN\SegAN-master\data_loader.py", line 83, in getitem
image = imread(self.imagePath)
File "E:\anaconda\envs\pytorch\lib\site-packages\tifffile\tifffile.py", line 715, in imread
with TiffFile(files, **kwargs_file) as tif:
File "E:\anaconda\envs\pytorch\lib\site-packages\tifffile\tifffile.py", line 2228, in init
fh = FileHandle(arg, mode='rb', name=name, offset=offset, size=size)
File "E:\anaconda\envs\pytorch\lib\site-packages\tifffile\tifffile.py", line 6031, in init
self.open()
File "E:\anaconda\envs\pytorch\lib\site-packages\tifffile\tifffile.py", line 6044, in open
self._fh = open(self._file, self._mode)
PermissionError: [Errno 13] Permission denied: 'D:\PycharmFile\ImageSegmentation\GAN\SegAN-master\preprocessed\train\sampleimages\volume'

Why does this error still occur when I use the lits dataset

[blublebee]

Thank you very much for sharing SegAN. Currently, I apply the SegAN to segment the rib. If I use the Unet with the resnet as the backbone, the dice can achieve 0.85. But for SegAN, I can only obtain dice of 0.7. Do you have any suggestion?
could you share your code for me？I can‘t run this code，the code can’t open .nii file.