search

richzhang / colorization

Automatic colorization using deep neural networks. "Colorful Image Colorization." In ECCV, 2016.
http://richzhang.github.io/colorization/
BSD 2-Clause "Simplified" License
3.34k stars 921 forks source link

Model architecture does not have H*W*Q size tensors for (a,b) probability #96

Open tathagatv opened 3 years ago

tathagatv commented 3 years ago

As mentioned in the paper, the loss function uses Z, Z^ which are of shape H*W*Q. However the model architecture computes the (a,b) probability distribution in a tensor of shape H/4 * W/4 * Q. How are we computing Z^ then? Is it not supposed to be predicted by the model and we use it for calculation of the loss?

simonebonato commented 2 years ago

We are also having problems because of this... If anyone could enlighten us it would be appreciated.