Datasets
Standard Dataset
IReNe: Instant Recoloring of Radiance Fields dataset
- Citation Author(s):
- Submitted by:
- Alessio Mazzucchelli
- Last updated:
- Tue, 10/15/2024 - 11:33
- DOI:
- 10.21227/r4h1-3638
- Data Format:
- Research Article Link:
- Links:
- License:
1 Views- Categories:
- Keywords:
Abstract
A dataset has been created by recoloring three existing datasets: NeRF Synthetic, LLFF, and Mip 360. The recoloring was performed to provide ground truth for validating recoloring applications. NeRF Synthetic was recolored using Blender, while LLFF and Mip 360 were processed in Photoshop. For each scene in the datasets, 11 images were recolored, ensuring consistency across the datasets. This dataset can serve as a benchmark for evaluating and comparing various recoloring methods, providing a reliable ground truth for validation in experiments involving NeRF-based and 3D scene reconstruction techniques.
All the names in the JSON are correct for the Train/Test division, don't count the validation one.
I dind't have time to recheck the poses since I was using them directly from COLMAP, so if you want to use the poses from JSON files please check if they are correctly related to the specified image.
When I have time I will fix the poses in the JSON as well. Sorry for the inconvenience.
Synthetic_new is the NeRF_synthetic rendered scenes without edited color, since in the original file some blender files were missing and I needed to have similar reflections materials for the recoloring
