By Jinjing Li, Chenghua Li, Fangya Li, Ruipeng Gang, Qian Zheng and Yuntian Cao

1. Dataset
2. Configuration
3. How to test
4. How to train
5. Visualization

Register a codalab account and log in, then find the download link on this page:

https://competitions.codalab.org/competitions/28161#participate-get-data

pip install -r requirements.txt

cd codes/models/dcnv2
python setup.py build develop  # build in your conda virtual environment

• download pretrained model Google Drive: https://drive.google.com/file/d/1wRsc9HQFcbT7nEFS50cfL9N3bql9cuGV/view?usp=sharing

• put this model 210000_G.pth in './experiments/Delete_0307/models',

• Prepare the data. Modify the dataroot_LDRs in ./codes/options/test/test_HDR.yml, then run:

cd codes
python test.py -opt options/test/test_HDR.yml

The test results will be saved to ./results/<your result name>.

• Prepare the data. Modify input_folder and save_folder in ./scripts/extract_subimgs_single.py, then run

cd scripts
python extract_subimgs_single.py

• Modify dataroot_LQ and dataroot_GT in ./codes/options/train/train_HDR.yml, then run:

cd codes
python train.py -opt options/train/train_HDR.yml

The models and training states will be saved to ./experiments/name.

• Because of storage limitations, we use (1, 6, 1060 // 4, 1900 // 4) as inputs. When submitting readme.txt, we use total_macs * 16 and mean_runtime * 16 as our result. You can try to set the scale=1 in calculate_ops_example.py if your GPU allowed.

• run:

python calculate_ops_example.py

In ./scripts, several scripts are available. data_io.py and metrics.py are provided by the competition organizer for reading/writing data and evaluation. Based on these codes, I provide a script for visualization by using the tone-mapping provided in metrics.py. Modify paths of the data in ./scripts/tonemapped_visualization.py and run

cd scripts
python tonemapped_visualization.py

to visualize the images.

The code is inspired by HDRUNet.