DREAMING - Diminished Reality for Emerging Applications in Medicine through Inpainting Dataset

Dataset for the DREAMING - Diminished Reality for Emerging Applications in Medicine through Inpainting Challenge

More information about the challenge and challenge registration can be found on the challenge platform!

Timeline:

8th January 2024: First subset of training & validation data available (scenes 0000-0009).

22nd January 2024: Second subset of training & validation data available (scenes 0010-0049).

26th January 2024: Full training & validation data available (scenes 0050-0100).

Description:

The dataset was created using Unreal Engine 5.1, Unreal MetaHumans, 3D-COSI surgical instruments, POV-Surgery grasp generation and EasySynth.

Each scene contains:

"color": RGB input images

"gt": RGB ground truth images

"mask": Mask defining the area to be inpainted. White -> Background, Black -> Inpainting area

"CameraPoses.csv": Camera poses

"CameraRig.json": Camera intrinsics

To cite the dataset, please use

@online{gsaxner_2024_10570773,
author = {Gsaxner, Christina and van Meegdenburg, Timo and Luijten, Gijs and Vu, Viet Duc and Puladi, Behrus and Egger, Jan},
title = {{DREAMING - Diminished Reality for Emerging Applications in Medicine through Inpainting Dataset}},
month = jan,
year = 2024,
publisher = {Zenodo},
doi = {10.5281/zenodo.10570773},
url = {https://doi.org/10.5281/zenodo.10570773}
}