Inference

[omagdy]

Inference

[omagdy]

Hi @dengyijia214

In any GAN model the inference is done using the generator after the training is complete. You can use the this function to automatically load the latest generator saved during training.

And then you can easily do inference with it like in this line of code. Make sure the input has 4 dimensions (number_of_images, x_size, y_szie, z_size).

Let me know if any part needs more clarification.

[dengyijia214]

Thanks for your reply, I got sick these days so I did not reply you on time. I have not describe my question clearly, I want to know if I want to get the whole predicted image, for example the whole 512512300 size CT, is there anything I could do?

[omagdy]

Hi @dengyijia214

You would have to split the whole image into smaller blocks using a package like this one and then group them back together.