-
Notifications
You must be signed in to change notification settings - Fork 3
Open
Description
After processing by running 'scripts/data_preparation/extract_subimages.py', totally obtained image number in 'datasets/DF2K/DF2K_train_LR_bicubic/X3_sub' is different from that in 'datasets/DF2K/DF2K_train_HR_sub'. So that error occurred when training launched. Yet, image numbers in 'X3_sub' and 'X4_sub' are same as that in 'HR_sub', after processing data . Are parameters below right?
LRx2 images
opt['input_folder'] = 'datasets/DF2K/DF2K_train_LR_bicubic/X2'
opt['save_folder'] = 'datasets/DF2K/DF2K_train_LR_bicubic/X2_sub'
opt['crop_size'] = 240
opt['step'] = 120
opt['thresh_size'] = 0
extract_subimages(opt)
LRx3 images
opt['input_folder'] = 'datasets/DF2K/DF2K_train_LR_bicubic/X3'
opt['save_folder'] = 'datasets/DF2K/DF2K_train_LR_bicubic/X3_sub'
opt['crop_size'] = 160
opt['step'] = 80
opt['thresh_size'] = 0
extract_subimages(opt)
LRx4 images
opt['input_folder'] = 'datasets/DF2K/DF2K_train_LR_bicubic/X4'
opt['save_folder'] = 'datasets/DF2K/DF2K_train_LR_bicubic/X4_sub'
opt['crop_size'] = 120
opt['step'] = 60
opt['thresh_size'] = 0
extract_subimages(opt)
Metadata
Metadata
Assignees
Labels
No labels