Replies: 1 comment 2 replies
-
What's your target platform ? Pytorch is a requirement here since we solve an online optimization process to get the transformation matrix between two given image views using automatic differentiation (Pytorch autograd). It's not a standard network that "generalizes, instead you need to "learn" every time the transformation. Thoughts @ducha-aiki ? |
Beta Was this translation helpful? Give feedback.
2 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
*Inference
As far as I know ImageRegistrator module is a little network that is trained using PyTorch every time when you use it.
Is there a way how I can use this in my network during inference time? Is it really neccesary to carry PyTorch and other stuff with it? I want to run it on the edge device.
Beta Was this translation helpful? Give feedback.
All reactions