HI, when I read your code, I meet some questions.
In your code, after registration by VTNAffine subnetwork, will the move image be warped with the Affine metric by the function Dense3DSpatialTransformer in spatial_transformer.py? And then will the moved image be send to the VTN dense deformable subnetwork to training further?
Is the Fast3DTransformer just only a version of a memory-efficient warping operation for Dense3DSpatialTransformer?
Thanks a lot!
HI, when I read your code, I meet some questions. In your code, after registration by VTNAffine subnetwork, will the move image be warped with the Affine metric by the function Dense3DSpatialTransformer in spatial_transformer.py? And then will the moved image be send to the VTN dense deformable subnetwork to training further? Is the Fast3DTransformer just only a version of a memory-efficient warping operation for Dense3DSpatialTransformer? Thanks a lot!