search

xmengli / H-DenseUNet

TMI 2018. H-DenseUNet: Hybrid Densely Connected UNet for Liver and Tumor Segmentation from CT Volumes
532 stars 161 forks source link

a question about upsampling when training model。 #50

Open shenhai1895 opened 5 years ago

shenhai1895 commented 5 years ago

i find that the ops up0_sum = add([line0, up0]) in denseunet.py is used in the training stage but not in the test stage in . is it right? and why to do it?

shenhai1895 commented 5 years ago

i can't understand that why you delete up0_sum = add([line0, up0]) in denseunet when it predicting.

shenhai1895 commented 5 years ago

i find this function load_weights_from_hdf5_group_by_name_mulgpu_twomodelcombine. so, when you train the denseunet2d, you use the unet connection, while you delete the unet connection when training the hdenseunet. do i got it? thank you for your reply.