complete script for reproducing results
#1
by
Guanxiong
- opened
Hi,
Thanks for sharing your nice work. It is very helpful for my research. I'm attempting to reproduce the results with this trained model, but I'm confronting some problems reproducing it correctly. Could you please give some hints about how to generate the mask and kspace from the fastmri dataset using your repository? I would appreciate it if there was a complete script for that.
Best,
Guanxiong Luo
guanxiong.luo@med.uni-goettingen.de
Guanxiong
changed discussion title from
complete script to r
to complete script
Guanxiong
changed discussion title from
complete script
to complete script for reproducing results
Hi @Guanxiong ,
Thanks so much for reaching out and sorry for replying only now, I had not seen your message.
I am not sure I understand the question:
- if you want to evaluate on test data, then the mask and kspace come directly from the fastmri data without any need to generate them.
- if you want to evaluate on validation data, the script here should help you do exactly that : https://github.com/zaccharieramzi/fastmri-reproducible-benchmark/blob/master/fastmri_recon/evaluate/scripts/xpdnet_eval.py
If you have any error using this script, please do not hesitate to file an issue on the github repo.
Cheers,
Zaccharie Ramzi