Benchmark for visual localization on 3D mesh models

Repository for the localization benchmark at v-pnk.github.io/cadloc. To upload the results to the benchmark, run the evaluation script on your pose estimates and create a pull request to add your results into the tables.

Citation

If you use the benchmark or the code from this repository, please cite the following paper:

@article{Panek2023cadloc,
title={{Visual Localization using Imperfect 3D Models from the Internet}},
author={Vojtech Panek and Zuzana Kukelova and Torsten Sattler},
booktitle={Proceedings of the IEEE/CVF conference on computer vision and pattern recognition},
year={2023}}

Installation

The evaluation tool and rendering script needs few Python packages:

conda create --name cadloc python=3.9
conda activate cadloc
pip3 install open3d==0.17 pycolmap==0.3 numpy argparse tqdm Pillow

Preparation of data for the benchmark

The evaluation script expects the estimated poses to be stored in the .txt format used by visuallocalization.net, i.e., each line contains an image name and estimated pose described by a quaternion q and translation vector t:

<name> <q[W]> <q[X]> <q[Y]> <q[Z]> <t[X]> <t[Y]> <t[Z]>

Uploading new benchmark results

Edit the table corresponding to your results and create a pull request.

License