search

LeeJunHyun / Image_Segmentation

Pytorch implementation of U-Net, R2U-Net, Attention U-Net, and Attention R2U-Net.
2.66k stars 594 forks source link

inputs of evaluation for training and validation #48

Closed CN-zdy closed 4 years ago

CN-zdy commented 4 years ago

@LeeJunHyun Thanks for sharing the code I found some differences in the evaluation of training and validation in solver.py. Is this correct? Because I am a beginner, so a little confused!

Train ### ### valid

SR = self.unet(images) SR = F.sigmoid(self.unet(images)) SR_probs = F.sigmoid(SR) ...... ...... acc += get_accuracy(SR,GT) acc += get_accuracy(SR,GT) SE += get_sensitivity(SR,GT) SE += get_sensitivity(SR,GT) SP += get_specificity(SR,GT) SP += get_specificity(SR,GT) PC += get_precision(SR,GT) PC += get_precision(SR,GT) ...... ......