Single Image Haze Removal Using AODNet in Pytorch
- Implementation of Boyi Li's Paper An All-in-One Network for Dehazing and Beyond on ICCV 2017.
- How to Use : download the whole project and run inference.py
- folder ./saved_models : where the trained models are saved, files are in .pth format.
- folder ./data/gt : groundtruth (haze free images) of the training data.
- folder ./data/hazy : corresponding hazy images of the training data.
- folder ./test_images : some testing images that appear in the original paper.
- data.py : function that loads the training data.
- train.py : train a new AODNet from scratch using training data saved in folder ./data/.
- model.py : definition of AODNet.
- utils.py : some auxiliary functions.
- inference.py : single image dehazing using the trained AODNet.
