Table 4 The results of average PSNR (dB) and SSIM on the Set5 [1], Set14 [37], and BSD200 [23]
From: Anchored neighborhood deep network for single-image super-resolution
Dataset | Scale | Bicubic | A+ | SRF | SRCNN | SCN | ANNet |
|---|---|---|---|---|---|---|---|
| Â | Â | PSNR/SSIM | PSNR/SSIM | PSNR/SSIM | PSNR/SSIM | PSNR/SSIM | PSNR/SSIM |
Set5 | 2 | 33.66/0.9299 | 36.55/0.9544 | 36.87/0.9556 | 36.34/0.9521 | 36.76/0.9545 | 36.73/0.9550 |
| Â | 3 | 30.39/0.9299 | 32.59/0.9088 | 32.71/0.9098 | 32.39/0.9033 | 33.04/0.9136 | 32.80/0.9117 |
| Â | 4 | 28.42/0.8104 | 30.28/0.8603 | 30.35/0.8600 | 30.09/0.8503 | 30.82/0.8728 | 30.38/0.8637 |
Set14 | 2 | 30.23/0.8687 | 32.28/0.9056 | 32.51/0.9074 | 32.18/0.9039 | 32.48/0.9067 | 32.43/0.9075 |
| Â | 3 | 27.54/0.7736 | 29.13/0.8188 | 29.23/0.8206 | 29.00/0.8145 | 29.37/0.8226 | 29.28/0.8217 |
| Â | 4 | 26.00/0.7019 | 27.32/0.7471 | 27.41/0.7497 | 27.20/0.7413 | 27.62/0.7571 | 27.40/0.7509 |
BSD200 | 2 | 29.70/0.8625 | 31.44/0.9031 | 31.65/0.9053 | 31.38/0.9287 | 31.63/0.9048 | 31.58/0.9052 |
| Â | 3 | 27.26/0.7638 | 28.36/0.8078 | 28.45/0.8095 | 28.28/0.8038 | 28.54/0.8119 | 28.46/0.8107 |
| Â | 4 | 25.97/0.6949 | 26.83/0.7359 | 26.89/0.7368 | 26.73/0.7291 | 27.02/0.7434 | 26.84/0.7365 |
Avg. | Â | 28.80/0.8151 | 30.53/0.8491 | 30.67/0.8505 | 30.39/0.8474 | 30.81/0.8542 | 30.77/0.8514 |