nadiabrancati / abnn-wsi-classification Goto Github PK

Thank you for sharing this work with us. I was reading your paper with interest. However, I came to some point in your paper that I didn't understand; how did you fine-tune your ResNet34 before using it for the feature extractor? you referred to some paper, but after investigating it I didn't get the idea!! Can you please explain the steps to do that? Thank you again

I tried testing the feature extractor with a Resnet-34 that has random weights, I tested it on a single WSI of 600mb, but I got an error stating that I need atleast 50GB of memory available.