search

lingyanruan / LaKDNet

Revisiting Image Deblurring with an Efficient ConvNet - An efficient CNN performs better than Transformer
GNU Affero General Public License v3.0
79 stars 2 forks source link

how to test model on our testing images #6

Open ghost opened 5 months ago

ghost commented 5 months ago

What is the script to deblur our own testing images ? If we want to evaluate model on our testing images then why we have to download the training dataset ?

fra888 commented 2 months ago

Hi I see nobody replied so I will tell you how I've done. I created a folder ./Test_sets/GOPRO/input/ and ./Test_sets/GOPRO/target/ then put your images in both folders, format must be png. Then run: python run.py --type Motion. I will modify the code for not testing anything, so it will need only one folder, but at the moment i use this trick for make it works.