Table 1.

The mean squared error between the high resolution ground truth (HRSI) and several methods are tabulated.

Method	Noise Level
	Low	Medium	High
Zero-Fill from 32x32	1.109	1.652	4.505
Bicubic from 32x32	2.794	3.129	3.820
16x16 D-UNet	1.863	2.420	2.761
24x24 D-UNet	1.139	1.316	1.745
32x32 D-UNet	0.7460	0.9722	1.599

These values are the total sum of the mean squared error over 169 test subjects. The 32x32 D-UNet reconstruction outperforms all of the other methods. With higher random noise present in the LRSI, the 16x16 and 24x24 D-UNets outperform both zero-filling and bicubic interpolation. It is important to note that this is true even though the zero-filling and bicubic interpolation methods are applied to a 32x32 image. Bold values indicate the method with the lowest mean squared error for each comparison.