Image Super Resolution On Bsd100 2X Upscaling

PSNR

評価結果

このベンチマークにおける各モデルのパフォーマンス結果

比較表

モデル名	PSNR
multi-level-wavelet-cnn-for-image-restoration	32.23
swinfir-revisiting-the-swinir-with-fast	32.64
hierarchical-information-flow-for-generalized	32.77
fast-accurate-and-lightweight-super-1	32.09
beyond-a-gaussian-denoiser-residual-learning	31.9
activating-more-pixels-in-image-super	32.74
wavemixsr-v2-enhancing-super-resolution-with	33.12
fast-accurate-and-lightweight-super	32.12
image-restoration-using-convolutional-auto	31.99
lightweight-image-super-resolution-with-1	32.19
channel-partitioned-windowed-attention-and	32.64
drct-saving-image-super-resolution-away-from	32.90
local-texture-estimator-for-implicit	32.44
swinfir-revisiting-the-swinir-with-fast	32.71
a-framework-for-real-time-object-detection	32.34
feedback-network-for-image-super-resolution	32.29
image-super-resolution-with-cross-scale-non	32.4
pre-trained-image-processing-transformer	32.48
accelerating-the-super-resolution	31.53
hmanet-hybrid-multi-axis-aggregation-network	32.79
densely-residual-laplacian-super-resolution	32.47
drct-saving-image-super-resolution-away-from	32.75
single-image-super-resolution-via-a-holistic	32.45
image-restoration-using-deep-regulated	31.86
lightweight-feature-fusion-network-for-single	31.96
channel-partitioned-windowed-attention-and	32.66
hierarchical-back-projection-network-for	32.33
deeply-recursive-convolutional-network-for	31.85
wavemixsr-a-resource-efficient-neural-network	33.08
activating-more-pixels-in-image-super	32.69