Open MarvinYeh opened 6 years ago
Thanks for the contribution. I'm wondering if your implementation should not rather go into a package for its own. What do you think?
Hi, I wasn't quite sure if I understand you correctly. Are you suggesting all the 3D-related features should be within an independent set of functions(or scripts?) from your own and leave all your original functions as it was?
Is there any official 3D unet code on tensorflow?
Hi,
I've updated the Unet3D and also create a demo on jupyter notebook. I don't have a proper 3d dataset to train on so currently it's trained using the generator, hence each slice of image is independent. it's still trainable but the result is not as good as using 2d Unet.