Ace

The issue proposed in [issue 4](https://github.com/felixlaumon/deform-conv/issues/4). I test the situation of : both network is training on the dataset (not deform mnist), the test accuracy ``` ---original net--- test_data: 99.03%...

mismatch of RayTransform in astra_cpu and astra_cuda

3

There is an "obvious" gap between RayTransform using `impl='astra_cuda'` and `impl='astra_cpu`. For example: ``` import numpy as np import odl size = 4 space = odl.uniform_discr([-size//2, -size//2], [size//2, size//2], [size,...

Question about phantom data

4

The value of ellipses is "unsimilar" with shepp logan? which is difficult to post-process methods. For example, the demo code as follow: ``` import odl from adler.odl.phantom import random_phantom from...

Question about [trunc_min, trunc_max] ?

1

Is there some reason to using [-160, 240] ? thank you.