Using the code for predictions on my dataset

[mjungiewicz-codete]

Thanks for the code!

I'd like to use a pretrained code on my own dataset. How to go about it? I tried to use the test_VGN.py file and I provided the paths to images, but the problem is that I don't have any labels for them and your code needs the labels...

[syshin1014]

Hi,
As you've already noticed, the current code was developed with public datasets that include GT labels, and it loads the GT label as well as an input image for evaluation. You need to modify the relevant part if you want to run it without GT labels.
Sorry for this inconvenience.

[Akhmetzhan]

Dear authors,
Thank you very much for sharing the code. I would like to use your model to get segmentated images from my own X ray coronary angiogramms. I am novel user of Python and PyTorch, that is why this task is quite difficult for me. I need segmentation for our scientific project. I tried to use this code : model = torch.load('/content/drive/MyDrive/Colab Notebooks/pretrained_model.tar.gz', map_location='cpu').
But there is an error - UnpicklingError: invalid load key, '\x1f'.
Could you please help me? I would be very grateful!