Research Article

Multilevel Feature Exploration Network for Image Superresolution

Table 3

Average PSNR/SSIM with degradation model BI, , and on five benchmarks. The best performances are shown in bold. The dash line means the paper does not report their performance.

ScaleModelSet5 [45]Set14 [46]B100 [47]Urban100 [48]Manga109 [49]
PSNR/SSIMPSNR/SSIMPSNR/SSIMPSNR/SSIMPSNR/SSIM

SRCNN [3]36.66/0.954232.42/0.906331.36/0.887929.50/0.894635.74/0.9661
FSRCNN [29]37.00/0.955832.63/0.908831.53/0.892029.88/0.902036.67/0.9694
VDSR [30]37.53/0.958733.03/0.912431.90/0.896030.76/0.914037.22/0.9729
DRCN [31]37.63/0.958833.04/0.911831.85/0.894230.75/0.913337.63/0.9723
LapSRN [9]37.52/0.959033.08/0.913031.80/0.895030.41/0.910037.27/0.9740
RAN [38]37.58/0.959233.10/0.913331.92/0.8963
DNCL [51]37.65/0.959933.18/0.914131.97/0.897130.89/0.9158
FilterNet [37]37.86/0.961033.34/0.915032.09/0.899031.24/0.9200
MRFN [39]37.98/0.961133.41/0.915932.14/0.899731.45/0.922138.29/0.9759
SeaNet-baseline [11]37.99/0.960733.60/0.917432.18/0.899532.08/0.927638.48/0.9768
DEGREE [13]37.58/0.958733.06/0.912331.80/0.8974
IRLP [40]37.83/0.969233.18/0.924732.01/0.902130.86/0.919237.71/0.9861
FSN [52]37.68/0.906533.51/0.918032.09/0.901531.68/0.9248
DSRLN [53]38.05/—33.56/—32.38/—31.79/—
MFSR38.07/0.960833.69/0.919132.22/0.900232.35/0.930738.75/0.9768

SRCNN [3]32.75/0.909029.28/0.820928.41/0.786326.24/0.798930.59/0.9107
FSRCNN [29]33.16/0.914029.43/0.824228.53/0.791026.43/0.808030.98/0.9212
VDSR [30]33.66/0.921329.77/0.831428.82/0.797627.14/0.827932.01/0.9310
DRCN [31]33.82/0.922629.76/0.831128.80/0.796327.15/0.827632.31/0.9328
DRRN [31]34.03/0.924429.96/0.834928.95/0.800427.53/0.837832.74/0.9390
RAN [38]33.71/0.922329.84/0.832628.84/0.7981
DNCL [51]33.95/0.923229.93/0.834028.91/0.799527.27/0.8326
FilterNet [37]34.08/0.925030.03/0.837028.95/0.803027.55/0.8380
MRFN [39]34.21/0.926730.03/0.836328.99/0.802927.53/0.838932.82/0.9396
SeaNet-baseline [11]34.36/0.928030.34/0.842829.09/0.805328.17/0.852733.40/0.9444
DEGREE [13]33.76/0.921129.82/0.832628.74/0.7950
IRLP [40]34.22/0.943730.16/0.847229.21/0.816327.58/0.832732.61/0.9492
DSRLN [53]34.56/—30.36/—29.29/—27.88/—
MFSR34.49/0.928030.42/0.844229.16/0.806828.39/0.857733.72/0.9457

SRCNN [3]30.48/0.862827.49/0.750326.90/0.710124.52/0.722127.66/0.8505
FSRCNN [29]30.71/0.865727.59/0.753526.98/0.715024.62/0.728027.90/0.8517
VDSR [30]31.35/0.883828.01/0.767427.29/0.725125.18/0.752428.83/0.8809
DRCN [31]31.53/0.885428.02/0.767027.23/0.723325.14/0.751028.98/0.8816
LapSRN [9]31.54/0.885028.19/0.772027.32/0.728025.21/0.756029.09/0.8845
RAN [38]31.43/0.884728.09/0.769127.31/0.7260
DNCL [51]31.66/0.887128.23/0.771727.39/0.728225.36/0.7606
FilterNet [37]31.74/0.890028.27/0.773027.39/0.729025.53/0.7680
MRFN [39]31.90/0.891628.31/0.774627.43/0.730925.46/0.765429.57/0.8962
SeaNet-baseline [11]32.18/0.894828.61/0.782227.57/0.735926.05/0.789630.44/0.9088
DEGREE [13]31.47/0.883728.10/0.766927.20/0.7216
IRLP [40]31.94/0.902128.49/0.791027.70/0.747925.61/0.776429.49/0.9025
FSN [52]32.10/0.895928.57/0.787427.53/0.743825.76/0.7817
MFSR32.26/0.896128.65/0.783827.63/0.738126.25/0.791930.62/0.9103