---
license: cc-by-nc-nd-4.0
---

Dataset for monocular visual relocalization on a COLMAP 3D reconstruction model.

This dataset was collected at EZOHUB Tokyo with the cooperation of SATUDORA HOLDINGS CO.,LTD.

### Dataset structure

- `train`         : Images used to construct the COLMAP model. 
- `model`         : The COLMAP model files.
- `test`          : Test images to run relocalization.
- `intrinsic.xml` : Intrinsic camera parameters.