search

kkjishnu / Image-Dehazing-with-GAN

A pytorch implementation of Cycle-Dehaze
38 stars 8 forks source link

readme

Image Dehaze with GAN

Dataset

NYU Depth dataset, containing 1449 pairs of hazy and ground truth images. The dataset can be downloaded and extracted by the following command. It will be placed into folders A and B.

wget -O data.mat http://horatio.cs.nyu.edu/mit/silberman/nyu_depth_v2/nyu_depth_v2_labeled.mat
python3 extract.py

Split the dataset into train and test and place them in a folder 

    |─ datasets                   
    |   |─ Haze2Dehaze        
    |   |   |─ train              # Training
    |   |   |   |─ A              
    |   |   |   |─ B              
    |   |   |─ test               # Testing
    |   |   |   |─ A              
    |   |   |   |─ B