Table 4. Real-number multi-scale training comparison. The red shows the best while the blue indicates the second-best performance.

Dataset	Testing scale\Trained scales	2, 3, 4 PSNR/SSIM	1.5, 2, 2.5, 3, 2.5, 4 PSNR/SSIM	1.5, 1.75, 2, 2.25, 2.5, 2.75, 3, 3.25, 3.5, 3.75, 4 PSNR/SSIM	VDSR [4, 28] (2, 3, 4) PSNR/SSIM
Set 5	1.5	40.40/0.9766	40.62/0.9771	40.59/0.9770	33.54/0.9503
	1.75	38.81/0.9677	38.81/0.9677	38.82/0.9677	35.91/0.9572
	2	37.52/0.9585	37.49/0.9584	37.49/0.9585	37.53/0.9587
	2.25	36.40/0.9496	36.34/0.9495	36.37/0.9495	35.12/0.9416
	2.5	35.39/0.9404	35.41/0.9406	35.45/0.9406	34.34/0.9325
	2.75	34.66/0.9326	34.63/0.9327	34.68/0.9328	33.54/0.9265
	3	33.85/0.9226	33.83/0.9226	33.86/0.9228	33.66/0.9213
	3.25	33.18/0.9137	33.13/0.9137	33.20/0.9138	32.47/0.9080
	3.5	32.64/0.9042	32.66/0.9045	32.69/0.9046	32.18/0.9006
	3.75	32.00/0.8957	31.99/0.8955	31.98/0.8953	31.57/0.8914
	4	31.48/0.8854	31.51/0.8852	31.50/0.8853	31.34/0.8838