search

AngeLouCN / CaraNet

Context Axial Reverse Attention Network for Small Medical Objects Segmentation
462 stars 30 forks source link

About the CaraNet.py #10

Closed 15104356116 closed 2 years ago

15104356116 commented 2 years ago

Thanks for your code! But I still can not figure what cfpnet_res2net_v4 () is. This code is on line 135 of CaraNet.py.

AngeLouCN commented 2 years ago

Hi, it should be the caranet(). However, it will not affect the use of code and I will fix it soon.

15104356116 commented 2 years ago

Sorry to bother you again. Here is my result. And I would like to know how can I optimize the results. Here are two sets of result, mask and image. Ty20180910_1_1 Ty20180910_1_1 Ty20180910_1_1 Ty20180910_3_7 Ty20180910_3_7 Ty20180910_3_7 .

AngeLouCN commented 2 years ago

For the CT slice, I think maybe you can consider the preprocessing, more data augmentation and enhancement. Also, you can try to choose only loss5 as loss function without deep supervising.

15104356116 commented 2 years ago

Thanks! I'll try it .🥰🥰