You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Specifically, it doesn't seem to give the same angles as the old Initial_classification when dealing with flipped/reflected images. To see this, consider the test script
just after the call to Initial_classification_FD brings down the MSE from about 0.65 to 0.50. The difference is even greater in less noisy situations, when the MSE for the class averages will actually be larger than those of the original images when not correcting for the angles.
Presumably this has something to do with how the sPCA coefficients and/or bispectrum is computed in the new version. That being said, adding the above line to the Initial_classification_FD should solve the problem.
The text was updated successfully, but these errors were encountered:
Per issue #31, Initial_classification_FD gives the wrong angles
for the rotated images (compared to what Initial_classification
gives and align_main expects). This is fixed by correcting those
angles.
Specifically, it doesn't seem to give the same angles as the old
Initial_classification
when dealing with flipped/reflected images. To see this, consider the test scriptNow adding the line
just after the call to
Initial_classification_FD
brings down the MSE from about 0.65 to 0.50. The difference is even greater in less noisy situations, when the MSE for the class averages will actually be larger than those of the original images when not correcting for the angles.Presumably this has something to do with how the sPCA coefficients and/or bispectrum is computed in the new version. That being said, adding the above line to the
Initial_classification_FD
should solve the problem.The text was updated successfully, but these errors were encountered: