3d-mri-brain-tumor-segmentation-using-autoencoder-regularization icon indicating copy to clipboard operation
3d-mri-brain-tumor-segmentation-using-autoencoder-regularization copied to clipboard

Published 20 hours ago • verified publisher icon IAmSuyogJadhav

Bilinear Upsampling

Open Yan107351111 opened this issue 4 years ago • 1 comments

In the paper, the author specifies using bilinear interpolation to upsample the features in the decoder heads. Going over your implementation, I saw you are using Kerase's UpSampling3D object to do that. But it seems to me that what this object does is repeat elements, not perform interpolation. Please, correct me if I'm wrong.

Yan107351111 avatar Jul 05 '20 10:07 Yan107351111

You are right. By default, the UpSampling2D layer in tensorflow will use a nearest neighbor algorithm to fill in the new rows and columns. This has the effect of simply doubling rows and columns.

FBehrad avatar Sep 19 '21 05:09 FBehrad