Datasets:
Tasks:
Image-to-Image
Modalities:
Image
Formats:
imagefolder
Languages:
English
Size:
10K - 100K
ArXiv:
License:
| Representative sample creation: | |
| This sample subset contains one complete scene selected from the MODEST dataset to provide reviewers with a representative view of dataset quality, organization, and optical variation. | |
| The selected scene preserves the original capture protocol, including: | |
| - stereo image pairs, | |
| - all focal length settings (28mm–70mm), | |
| - all aperture settings (f/2.8–f/22), | |
| - calibration assets, | |
| - rectified outputs. | |
| The sample was chosen by selecting one full scene from the ten captured scenes without modifying image resolution, metadata, or preprocessing, thereby preserving the original dataset fidelity and folder structure. | |
| This subset is intended only for qualitative inspection and structural verification; the full dataset (20,000 images, ~400GB) is available through the main dataset URL. | |