Research Article

Lightweight Image Super-Resolution Based on Re-Parameterization and Self-Calibrated Convolution

Table 2

Average PSNR (dB)/SSIM values with the scale factors ×2, ×3, and ×4 on Set5, Set14, BSD100, and Urban100. The best performance is highlighted in red, and the second-best performance is highlighted in blue.

MethodScaleParamsSet5 PSNR/SSIMSet14 PSNR/SSIMBSD100 PSNR/SSIMUrban100 PSNR/SSIM

SRCNN [12]×28K36.66/0.954232.45/0.906731.36/0.887929.50/0.8946
FSRCNN [13]×213K37.00/0.955832.63/0.908831.53/0.892029.88/0.9020
VDSR [14]×2666K37.53/0.958733.03/0.912431.90/0.896030.76/0.9140
DRCN [16]×21774K37.63/0.958833.04/0.911831.85/0.894230.75/0.9133
LapSRN [24]×2251K37.52/0.959132.99/0.912431.80/0.895230.41/0.9103
DRRN [17]×2298K37.74/0.959133.23/0.913632.05/0.897331.23/0.9188
MemNet [25]×2678K37.78/0.959733.28/0.914232.08/0.897831.31/0.9195
IDN [20]×2553K37.83/0.960033.30/0.914832.08/0.898531.27/0.9196
EDSR-baseline [15]×21370K37.99/0.960433.57/0.917532.16/0.899431.98/0.9272
SRMDNF [46]×21511K37.79/0.960133.32/0.915932.05/0.898531.33/0.9204
CARN [19]×21592K37.76/0.959033.52/0.916632.09/0.897831.92/0.9256
IMDN [27]×2694K38.00/0.960533.63/0.917732.19/0.899632.17/0.9283
MADNet [47]×2878K37.94/0.960433.46/0.916732.10/0.898831.74/0.9246
MSICF [48]×24292K37.89/0.960533.41/0.915332.15/0.899231.47/0.9220
MSWSR [49]×21228K37.49/0.958333.23/0.912331.88/0.892931.14/0.9169
RepSCN(Ours)×2753K38.01/0.960633.70/0.919232.19/0.899932.36/0.9307
SRCNN [12]×38K32.75/0.909029.30/0.821528.41/0.786326.24/0.7989
FSRCNN [13]×313K33.18/0.914029.37/0.824028.53/0.791026.43/0.8080
VDSR [14]×3666K33.66/0.921329.77/0.831428.82/0.797627.14/0.8279
DRCN [16]×31774K33.82/0.922629.76/0.831128.80/0.796327.15/0.8276
LapSRN [24]×3502K33.81/0.922029.79/0.832528.82/0.798027.07/0.8275
DRRN [17]×3298K34.03/0.924429.96/0.834928.95/0.800427.53/0.8378
MemNet [25]×3678K34.09/0.924830.00/0.835028.96/0.800127.56/0.8376
IDN [20]×3553K34.11/0.925329.99/0.835428.95/0.801327.42/0.8359
EDSR-baseline [15]×31555K34.37/0.927030.28/0.841729.09/0.805228.15/0.8527
SRMDNF [46]×31528K34.12/0.925430.04/0.838228.97/0.802527.57/0.8398
CARN [19]×31592K34.29/0.925530.29/0.840729.06/0.803428.06/0.8493
IMDN [27]×3703K34.36/0.927030.32/0.841729.09/0.804628.17/0.8519
MADNet [47]×3930K34.26/0.926230.29/0.841029.04/0.803327.91/0.8464
MSICF [48]×34292K34.24/0.926630.09/0.837129.01/0.802427.69/0.8411
MSWSR [49]×3−/−−/−−/−−/−
RepSCN(Ours)×3761K34.49/0.927730.38/0.843329.09/0.805428.30/0.8553
SRCNN [12]×48K30.48/0.862827.50/0.751326.90/0.710124.52/0.7221
FSRCNN [13]×413K30.72/0.866027.61/0.755026.98/0.715024.62/0.7280
VDSR [14]×4666K31.35/0.883828.01/0.767427.29/0.725125.18/0.7524
DRCN [16]×41774K31.53/0.885428.02/0.767027.23/0.723325.14/0.7510
LapSRN [24]×4502K31.54/0.885228.09/0.770027.32/0.727525.21/0.7562
DRRN [17]×4298K31.68/0.888828.21/0.772027.38/0.728425.44/0.7638
MemNet [25]×4678K31.74/0.889328.26/0.772327.40/0.728125.50/0.7630
IDN [20]×4553K31.82/0.890328.25/0.773027.41/0.729725.41/0.7632
EDSR-baseline [15]×41518K32.09/0.893828.58/0.781327.57/0.735726.04/0.7849
SRMDNF [46]×41552K31.96/0.892528.35/0.778727.49/0.733725.68/0.7731
CARN [19]×41592K32.13/0.893728.60/0.780627.58/0.734926.07/0.7837
IMDN [27]×4715K32.21/0.894828.58/0.781127.56/0.735326.04/0.7838
MADNet [47]×41002K32.11/0.893928.52/0.779927.52/0.734025.89/0.7782
MSICF [48]×44292K31.91/0.892328.35/0.775127.46/0.730825.64/0.7692
MSWSR [49]×41228K32.01/0.891428.47/0.777627.48/0.731125.78/0.7744
RepSCN(Ours)×4772K32.22/0.895328.61/0.782027.58/0.736326.17/0.7881