Access the Challenge Dataset
Training Data:
[Baidu Netdisk]
[OneDrive]
Validation Data(input):
[Baidu Netdisk]
[OneDrive]
Testing Data(input):
[TBD]
[TBD]
The Dataset is for this chanllenge and research purpose only. Commercial use is not permitted.
Content of The Downloaded Package
In the dataset, you will find a folder for training data and a folder for testing data.
Train
In the training dataset folder, participants will find 300 high-resolution training images, each with dimensions of 4480x6720. These images are formatted as '.mat' files, which can be accessed using scipy.io. Each file contains the input image in 14-bit Raw format, the corresponding 20-bit HDR image, and additional image profiles including white balance and color correction matrix.
Validation
The validation dataset consists of 30 pairs of images, each cropped to 1024x1024 size with high dynamic range features. The format of the validation images is consistent with the training set, stored in '.mat' files. These files include input raw images, output HDR images, and image profiles.
Test
For the test dataset, there are 30 raw images provided as input, each with a spatial resolution of 1024x1024. Similar to the other datasets, these images are available in '.mat' format, including input raw images and image profiles. However, the test dataset does not include the ground truth HDR images.
Citation
If you use this dataset for any purpose, please cite our previous work