Closed dyang37 closed 2 months ago
I tried to run the test scripts, but got an error.
It seems that you are running the program on the front end of Gilbreth (gilbreth-fe00).
You would need to reserve a computation node:
sinteractive -A standby --time 4:00:00 -n 64 -N 1 --gpus-per-node=1
Then you may run the program on the node that you reserved.
This ran fine, so I'm going to accept the PR to prerelease. We can discuss it tomorrow and see if it is ready for integration into the main branch.
This PR contains the function to generate the weight for metal artifact reduction, along with a function to segment an image with Otsu's method.
Function interfaces: MAR weight:
Otsu's method
Both the MAR weight and Otsu's method have been tested on two datasets (MAR dataset, AutoInjector).
To test:
mar_weight
, and build mbirjax.mbirjax_applications
and checkout branchmar_weight
.demo_nsi_mar.py
. This script takes about 5 minutes to run on Gilbreth.