-
2.2.1の式7以降が地味に重い
Primal-dual approach(線型計画法に登場する手法らしい)というのが出てくるのだが詳しい説明もなく参考文献もかなり重い
```
As a function of ξ, the convex sum g(u)k∇ξ(u)k+ Q(u) is a small modification of
the TV-L ROF image denoi…
-
We consider TV-like problem of the form
&space;+&space;g(D&space;x))
where $f,g$ ar…
-
I have some problem about calculate ||K||, could you give some advice?
1. the image vector initialize by sinogram back projection, is it necessary? I just initialize by rand number.
2. the differen…
cgllv updated
6 years ago
-
Why do you generate new training data pairs just every 10th iteration? For instance [here](https://github.com/adler-j/learned_primal_dual/blob/master/ellipses/learned_primal_dual.py#L162).
Also [t…
-
I maintain a tomography processing package (https://gitlab.com/aaherzing/tomotools) and I'm interested in integrating your C-P algorithm to my reconstruction module to add a TV minimization option. D…
-
Hi,
I had one issue when running the code in ellipses :
File "C:\Users\ZG\AppData\Local\Continuum\anaconda3\lib\site-packages\tensorflow\python\ops\variable_scope.py", line 848, in _get_singl…
-
I just downloaded the repository, and it was a staggering 63MB. A quick run with [this script](https://gist.github.com/adler-j/95f69f24d7fe2dd3772a562660a138a2) gave this output:
```bash
All siz…
-
I've forked pygsp and started implementing the tikhonov optimization on my branch, but it appears that
1) The current implementation of `optimization.prox_tv` does not work because
2) `pyunlocbox` …
-
With reload modules on in Spyder, we can get bugs where `isinstance(RealNumbers(), RealNumbers)` does not seem to hold, at least if the parameters are given in a more complicated manner.
-
https://github.com/odlgroup/odl/blob/6a18fb0793c8ebe236b7acea835e62fbce20d08d/odl/solvers/nonsmooth/forward_backward.py#L19
I just noticed that the `forward_backward_pd` is pretty much the same alg…