Image Super Resolution On Set14 2X Upscaling

评估指标

PSNR

SSIM

评测结果

各个模型在此基准测试上的表现结果

比较表格

模型名称	PSNR	SSIM
drct-saving-image-super-resolution-away-from	35.36	0.9302
ml-craist-multi-scale-low-high-frequency	33.64	0.9213
channel-partitioned-windowed-attention-and	34.91	0.9277
a-framework-for-real-time-object-detection	33.97	0.922
multi-level-wavelet-cnn-for-image-restoration	33.7	-
mair-a-locality-and-continuity-preserving	34.75	0.9268
cascade-convolutional-neural-network-for	34.34	0.9240
activating-more-pixels-in-image-super	35.29	0.9293
activating-more-pixels-in-image-super	35.13	0.9282
channel-partitioned-windowed-attention-and	34.97	0.9280
drct-saving-image-super-resolution-away-from	34.96	0.9287
densely-residual-laplacian-super-resolution	34.43	0.9247
hierarchical-back-projection-network-for	33.78	0.921
deeply-recursive-convolutional-network-for	33.04	-
beyond-a-gaussian-denoiser-residual-learning	33.03	-
deep-back-projection-networks-for-single	34.09	0.921
learning-deep-cnn-denoiser-prior-for-image	30.79	-
swinfir-revisiting-the-swinir-with-fast	35.17	-
ml-craist-multi-scale-low-high-frequency	33.77	0.922
single-image-super-resolution-via-a-holistic	34.24	0.9224
accurate-image-super-resolution-using-very	33.03	-
hierarchical-information-flow-for-generalized	35.27	0.9311
fast-and-accurate-image-super-resolution-by	33.05	.9126
hmanet-hybrid-multi-axis-aggregation-network	35.33	0.9297
lightweight-image-super-resolution-with-1	33.63	-
swinfir-revisiting-the-swinir-with-fast	34.93	0.9276
image-super-resolution-with-cross-scale-non	34.12	0.9223
feedback-network-for-image-super-resolution	33.82	-
sub-pixel-back-projection-network-for	33.62	0.9178
fast-and-accurate-image-super-resolution-by	32.71	.9090
fast-accurate-and-lightweight-super	33.55	-
image-restoration-using-convolutional-auto	32.94	0.9144
fast-accurate-and-lightweight-super-1	33.52	-
fast-accurate-and-lightweight-super-1	33.26	-
local-texture-estimator-for-implicit	34.25	-