Skip to main content
. 2021 Jul 26;37(8):085006. doi: 10.1088/1361-6420/ac104f

Figure 7.

Figure 7.

A comparison of the performance of equivariant and ordinary learned proximal gradient methods trained on training sets of various sizes for the CT reconstruction problem. The methods are tested on images that have not been seen during training time, both in the same orientations as were observed during training (‘upright test images’) and rotated at random angles (‘rotated test images’). The performance is evaluated on the lung regions only.