Image Super Resolution On Urban100 2X

평가 지표

PSNR

SSIM

평가 결과

이 벤치마크에서 각 모델의 성능 결과

모델 이름	PSNR	SSIM	Paper Title	Repository
SwinOIR	32.83	0.9353	Resolution Enhancement Processing on Low Quality Images Using Swin Transformer Based on Interval Dense Connection Strategy
HMA†	35.24	0.9513	HMANet: Hybrid Multi-Axis Aggregation Network for Image Super-Resolution
SPBP-L+	32.07	0.9277	Sub-Pixel Back-Projection Network For Lightweight Single Image Super-Resolution
FACD	32.878	-	Feature-domain Adaptive Contrastive Distillation for Efficient Single Image Super-Resolution	-
CSNLN	33.25	0.9386	Image Super-Resolution with Cross-Scale Non-Local Attention and Exhaustive Self-Exemplars Mining
MWCNN	32.3	-	Multi-level Wavelet-CNN for Image Restoration
VDSR [[Kim et al.2016a]]	30.76	-	Accurate Image Super-Resolution Using Very Deep Convolutional Networks
HAT	34.81	0.9489	Activating More Pixels in Image Super-Resolution Transformer
DRCT-L	35.17	0.9516	DRCT: Saving Image Super-resolution away from Information Bottleneck
DBPN-RES-MR64-3	32.92	0.935	Deep Back-Projection Networks for Single Image Super-resolution
PMRN+	32.78	0.9342	Sequential Hierarchical Learning with Distribution Transformation for Image Super-Resolution	-
DRLN+	33.54	0.9402	Densely Residual Laplacian Super-Resolution
HAT-L	35.09	0.9505	Activating More Pixels in Image Super-Resolution Transformer
ML-CrAIST-Li	32.93	0.9361	ML-CrAIST: Multi-scale Low-high Frequency Information-based Cross black Attention with Image Super-resolving Transformer
DRCN [[Kim et al.2016b]]	30.75	-	Deeply-Recursive Convolutional Network for Image Super-Resolution
DnCNN-3	30.74	-	Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising
SwinFIR	34.57	0.9473	SwinFIR: Revisiting the SwinIR with Fast Fourier Convolution and Improved Training for Image Super-Resolution
Hi-IR-L	35.16	0.9505	Hierarchical Information Flow for Generalized Efficient Image Restoration	-
ML-CrAIST	33.04	0.937	ML-CrAIST: Multi-scale Low-high Frequency Information-based Cross black Attention with Image Super-resolving Transformer
DRCT	34.54	0.9474	DRCT: Saving Image Super-resolution away from Information Bottleneck

0 of 29 row(s) selected.