search

wasidennis / AdaptSegNet

Learning to Adapt Structured Output Space for Semantic Segmentation, CVPR 2018 (spotlight)
850 stars 203 forks source link

classifier module #73

Closed jianingwangind closed 4 years ago

jianingwangind commented 5 years ago

I see that in model/deeplab_multi.py, the forward function of the Classifier_Module. The output seems to only the sum of the outputs from the first two dilated convolution layers. Is it so designed?

wasidennis commented 4 years ago

Please check this previous issue.

https://github.com/wasidennis/AdaptSegNet/issues/4