Open JevgenijaAksjonova opened 2 years ago
Sorry, my code is a mess, give a bit more days to do further refactorization, also, the experiments are not doing a joint model for wavefront set extraction, but rather wavefront set inpainting. You could start by training the U-Net in the Wavefront set inpainting folder.
Hi, I have uploaded a notebook where I gather all the code I used for the joint reconstruction. I have not fully tested the notebook yet but you can already take a look I will be updating it in the next few days. But the idea of the skeleton of the whole pipeline is already there.
Hi and thank you for sharing the code. As I understand from the description the code corresponding to the experiments in the paper should be in the folder Joint_CT_WFset_inpaint. I see there are notebooks that implement training the benchmarks, however I don't see a file that implements training of the joint model for reconstruction and WFset extraction.