SR Datasets

有挺多人问我关于超分辨率重建的数据集的问题，这里把超分辨率重建领域常用的数据集list出来

Name	Datasets	Short Description	Google Drive	Baidu Drive
Classical SR Training	T91	_{91 images for training}	Google Drive	Baidu Drive
	BSDS200	_{A subset (train) of BSD500 for training}
	General100	_{100 images for training}
Classical SR Testing	Set5	_{Set5 test dataset}
	Set14	_{Set14 test dataset}
	BSDS100	_{A subset (test) of BSD500 for testing}
	urban100	_{100 building images for testing (regular structures)}
	manga109	_{109 images of Japanese manga for testing}
	historical	_{10 gray LR images without the ground-truth}
2K Resolution	DIV2K	_{proposed in NTIRE17(800 train and 100 validation)}	Google Drive	Baidu Drive
	Flickr2K	_{2650 2K images from Flickr for training}
	DF2K	_{A merged training dataset of DIV2K and Flickr2K}
OST (Outdoor Scenes)	OST Training	_{7 categories images with rich textures}	Google Drive	Baidu Drive
OST (Outdoor Scenes)	OST300	_{300 test images of outdoor scences}	Google Drive	Baidu Drive
PIRM	PIRM	_{PIRM self-val, val, test datasets}	Google Drive	Baidu Drive

Reference

https://github.com/xinntao/BasicSR