Alan Morris
Alan Morris
The versions of VTK, ITK (perhaps others too) do not build with GCC 13. https://github.com/InsightSoftwareConsortium/ITK/issues/4607 https://gitlab.kitware.com/vtk/vtk/-/issues/18782
Would like to be able to copy these values
It looks like there may be an intermittent issue when re-running testing on a DeepSSM project in studio. I think the python side may be writing files at the same...
Add the ability to chart generalization, compactness and specificity for shape, shape+scalar, and scalar alone.
E.g. Set to 100%: ``` [2024-02-13 16:45:21.889] [error] Exception Caught: IndexError: index 0 is out of bounds for axis 0 with size 0 At: /Users/amorris/sci/shapeworks/code/Python/DataAugmentationUtilsPackage/DataAugmentationUtils/Embedder.py(46): run_PCA /Users/amorris/sci/shapeworks/code/Python/DataAugmentationUtilsPackage/DataAugmentationUtils/Embedder.py(25): __init__ /Users/amorris/sci/shapeworks/code/Python/DataAugmentationUtilsPackage/DataAugmentationUtils/DataAugmentation.py(36): point_based_aug...
Something is wrong with the scalar field particle assignment. This is correct: But this for normals is wrong: Resulting in reconstruction like this:
The DeepSSM functionality (in studio, and python) needs some input verification. Example issues: * Unaligned shape/image (we can at least check if the shape is inside the image) * Resulting...
If you run your new model using fixed domains and specify a different number of particles than the original particle count, not great things happen: https://shapeworks.discourse.group/t/new-shape-fixed-domains-c/202 It would be good...
Currently, at least when Procrustes is on, the world particles of fixed domains are moving. Likely because they are being moved by procrustes.