Are you sincere?
morizin opened this issue · 3 comments
ori_image: (H, W) numpy array of gray image
logit_map: (H, W) numpy array of model prediction
prob_map: (H, W) numpy array of model prediction with prob
is the size of the Logit maps is (H,W). when i train other network and removed the global pool layer and final linear layer then also i get the size of (N, C , 20,20) why is this like this when i print the shape of self.model.features or is this true
ori_image: (H, W) numpy array of gray image logit_map: (H, W) numpy array of model prediction prob_map: (H, W) numpy array of model prediction with probis the size of the Logit maps is (H,W). when i train other network and removed the global pool layer and final linear layer then also i get the size of (N, C , 20,20) why is this like this when i print the shape of self.model.features or is this true
Hi,
Thanks for your interest in our project.
From your comment, i guess your question is what the shape of the logit_map is.
To your questions ,the shape of logit maps is (H, W), but H & W here are not the same as the (H, W) of the ori_image. The actual size of logit maps is downsampled by several times from the original image size. So that's why you got (N, C , 20,20) output tensor size at the last Conv layer.
@morizin
Thanks for your reply. but while we are plotting heatmap
if in my case: for each disease
original_image (512,512)
logit_map (20,20)
prob_map (20,20)
will plot the heatmap