Skip to content

Fix device mismatch in SAM preprocessing #78

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.

Sign up for GitHub

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

Open
danikvh wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Fix device mismatch in SAM preprocessing #78

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions cellSAM/sam_inference.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -120,9 +120,9 @@ def predict_transforms(self, imgs):

return imgs

def sam_preprocess(self, x: torch.Tensor, return_paddings=False):
def sam_preprocess(self, x: torch.Tensor, return_paddings=False, device=None):
"""Normalize pixel values and pad to a square input."""
x = (x - self.model.pixel_mean) / self.model.pixel_std
x = (x - self.model.pixel_mean.to(device)) / self.model.pixel_std.to(device)

h, w = x.shape[-2:]
padh = self.model.image_encoder.img_size - h
Expand All @@ -140,7 +140,7 @@ def sam_bbox_preprocessing(self, x, device=None):
x = [
torch.from_numpy(img).permute(2, 0, 1).contiguous().to(device) for img in x
]
x = [self.sam_preprocess(img, return_paddings=True) for img in x]
x = [self.sam_preprocess(img, return_paddings=True, device=device) for img in x]
x, paddings = zip(*x)
preprocessed_img = torch.stack(x, dim=0)
return preprocessed_img, paddings
Expand Down