search

Huang-ShiRui / Semi-UIR

Official code for "Contrastive Semi-supervised Learning for Underwater Image Restoration via Reliable Bank" (2023 CVPR)
113 stars 10 forks source link

How to separete the dataset into testS and testR and the validation dataset in details? #22

Open wwxb2012 opened 10 months ago

wwxb2012 commented 10 months ago

Can you provide the link for the training set and testing dataset? As the author mentioned in the article, UIEB dataset has 890 images, 800 are in the training dataset and 90 are in the testing dataset. However, according to the Supplementary Material, the validation set has 80 images from UIEB. Meanwhile, UWCNN dataset has 800 images in the training set, 110 images in the testing set, 120 images in the validation dataset while it contains 1449 images in total. It's confusing about how to divide the dataset.

tengxiao14 commented 4 months ago

Do you know how to split the dataset? I have the same issue