-
Notifications
You must be signed in to change notification settings - Fork 3
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add mirroring test-time augmentation #96
base: main
Are you sure you want to change the base?
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Tested and working fine, thanks!
Co-authored-by: Jan Valosek <[email protected]>
sw_batch_size=4, predictor=net, overlap=0.5, progress=False)[0], dims=[axis] | ||
) | ||
# average the prediction | ||
batch["pred"] /= 3 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Just wondering if averaging should be done after clamping ?
Because averaging is done before clamping, if a voxel is seen only once, its value will be 0.33 and therefore will be removed afterward at clamping (done at 0.5).
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
oh man, this is an excellent catch! thanks for pointing this much important thing out! I'm going to fix it in the next commit!
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Fixed in commit 4beb690
@plbenveniste do you think it's good to merge after the update of moving up |
Looks good to me :) |
This is a small PR that updates the inference script by adding mirroring test-time augmentation. Mirroing is also done in nnunet inference and @plbenveniste also noticed improved predictions for canproco SC seg preds.