Image Super Resolution On Div2K Val 4X

평가 지표

LPIPS

PSNR

SSIM

평가 결과

이 벤치마크에서 각 모델의 성능 결과

모델 이름	LPIPS	PSNR	SSIM	Paper Title
LINF	0.112	27.33	0.76	Local Implicit Normalizing Flow for Arbitrary-Scale Image Super-Resolution
EDSR	-	29.25	0.9017	Enhanced Deep Residual Networks for Single Image Super-Resolution
RankSRGAN	-	26.55	0.75	Implicit Diffusion Models for Continuous Super-Resolution
GOUB	0.220	26.89	0.7478	Image Restoration Through Generalized Ornstein-Uhlenbeck Bridge
FxSR-PD t=0.8	0.1028	27.51	0.789	Flexible Style Image Super-Resolution using Conditional Objective
LAR-SR	-	27.03	0.77	Implicit Diffusion Models for Continuous Super-Resolution
Bicubic	-	26.7	0.77	Implicit Diffusion Models for Continuous Super-Resolution
LINF t=0.0	0.248	29.14	0.83	Local Implicit Normalizing Flow for Arbitrary-Scale Image Super-Resolution
HCFlow	-	27.02	0.76	Implicit Diffusion Models for Continuous Super-Resolution
LIIF	-	29	0.89	Learning Continuous Image Representation with Local Implicit Image Function
HCFlow++	-	26.61	0.74	Implicit Diffusion Models for Continuous Super-Resolution
SRFlow-LP	0.109	27.51	0.78	Boosting Flow-based Generative Super-Resolution Models via Learned Prior
ESRGAN	-	26.22	0.75	Implicit Diffusion Models for Continuous Super-Resolution
PixelRL-SR	-	28.08	0.8140	Multi-Step Reinforcement Learning for Single Image Super-Resolution
LINF-LP	0.105	28.00	0.78	Boosting Flow-based Generative Super-Resolution Models via Learned Prior
AESOP	0.0893	29.137	0.8023	Auto-Encoded Supervision for Perceptual Image Super-Resolution
RCOT	0.104	28.41	0.804	Residual-Conditioned Optimal Transport: Towards Structure-Preserving Unpaired and Paired Image Restoration
SRFlow	0.12	27.09	0.76	SRFlow: Learning the Super-Resolution Space with Normalizing Flow
IDM	-	27.59	0.78	Implicit Diffusion Models for Continuous Super-Resolution
FxSR-PD t=0.0	0.239	29.24	0.8383	Flexible Style Image Super-Resolution using Conditional Objective

0 of 21 row(s) selected.