Image Super Resolution On Manga109 4X

Metrics

PSNR

SSIM

Results

Performance results of various models on this benchmark

Model Name	PSNR	SSIM	Paper Title	Repository
MemNet	29.42	0.8942	MemNet: A Persistent Memory Network for Image Restoration
ML-CrAIST	31.17	0.9176	ML-CrAIST: Multi-scale Low-high Frequency Information-based Cross black Attention with Image Super-resolving Transformer
bicubic	24.89	0.7866	ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks
SRGAN + Residual-in-Residual Dense Block	31.66	0.9196	ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks
CARN	30.40	0.9082	Fast, Accurate, and Lightweight Super-Resolution with Cascading Residual Network
DBPN-RES-MR64-3	31.74	0.921	Deep Back-Projection Networks for Single Image Super-resolution
MaIR+	32.66	0.9297	MaIR: A Locality- and Continuity-Preserving Mamba for Image Restoration	-
HAT_FIR	33.03	-	SwinFIR: Revisiting the SwinIR with Fast Fourier Convolution and Improved Training for Image Super-Resolution
SwinIR	32.22	0.9273	SwinIR: Image Restoration Using Swin Transformer
DAT+	32.67	0.9301	Dual Aggregation Transformer for Image Super-Resolution
SAN	31.66	0.9222	Second-Order Attention Network for Single Image Super-Resolution	-
+SPN	31.75	0.9229	Learning from History: Task-agnostic Model Contrastive Learning for Image Restoration
RDN	31.0	0.9151	Residual Dense Network for Image Super-Resolution
ABPN	31.79	0.921	Image Super-Resolution via Attention based Back Projection Networks
RGT+	32.68	0.9303	Recursive Generalization Transformer for Image Super-Resolution
HMA†	33.19	0.9344	HMANet: Hybrid Multi-Axis Aggregation Network for Image Super-Resolution
GMFN	31.24	0.9174	Gated Multiple Feedback Network for Image Super-Resolution
SRCNN	27.58	0.8555	Image Super-Resolution Using Deep Convolutional Networks
RCAN	31.22	0.9173	Image Super-Resolution Using Very Deep Residual Channel Attention Networks
DRLN+	31.78	0.9211	Densely Residual Laplacian Super-Resolution

0 of 49 row(s) selected.