shervinmin/deepcovid

deepcovid's Issues

Val file is not included in the Dataset provided

FileNotFoundError: [WinError 3] The system cannot find the path specified: './data/val'

Since Val folder is not included in the provided dataset, the program practically fails to go ahead from the building stage.
Do update the dataset with a valid Val folder.

FileNotFoundError: [Errno 2] No such file or directory: './data/val'

The dataset does not have a validation split, however the training code in ResNet18_train.py uses it.
Please share the validation split so that the results can be reproduced.
Following code
$ python ResNet18_train.py --dataset_path ./data/ --batch_size 20 --epoch 50 --num_workers 4 --learning_rate 0.001

gives following error.

Traceback (most recent call last): File "ResNet18_train.py", line 59, in <module> image_datasets = {x: datasets.ImageFolder(os.path.join(data_dir, x), File "ResNet18_train.py", line 59, in <dictcomp> image_datasets = {x: datasets.ImageFolder(os.path.join(data_dir, x), File "/Library/Python/3.8/site-packages/torchvision/datasets/folder.py", line 226, in __init__ super(ImageFolder, self).__init__(root, loader, IMG_EXTENSIONS if is_valid_file is None else None, File "/Library/Python/3.8/site-packages/torchvision/datasets/folder.py", line 108, in __init__ classes, class_to_idx = self._find_classes(self.root) File "/Library/Python/3.8/site-packages/torchvision/datasets/folder.py", line 137, in _find_classes classes = [d.name for d in os.scandir(dir) if d.is_dir()] FileNotFoundError: [Errno 2] No such file or directory: './data/val'

about

Hi Shervin,

Could you have update the code about squeezeNet and DenseNet, and the inference.py didn't run the right results, sad....

With Regards

AssertionError - inference.py

Number of manager items must equal union of block items

manager items: 2, # tot_items: 0

How to solve this error?

Inference predicting everything as non-covid

Hi Shervin,

I have tried to execute the same code, but the code is predicting every image as a non-covid one, when I test on the test-covid images. Could you please help me to resolve the issue.

With Regards,
Aparna Reji

Pre-trained models?

Could you please share some pre-trained networks, to test the algorithms directly? Thanks.

Dataset Incomplete

Hi. The number of images per category in the provided link (to download the dataset) doesn't match with the statistics in table 1 of your paper.

For the training set, Non-COVID images should be 3000 but in the provided link it is 580 images. Please upload the complete dataset.