Alejandro Fontan · Javier Civera · Michael Milford

Any-Feature V-SLAM is an automated visual SLAM library for Monocular cameras capable of switching to a chosen type of feature effortlessly and without manual intervention.

AnyFeature-VSLAM runs seamlessly as a baseline for experiments within the VSLAM-LAB framework.

Note: More updates from the publication RSS 2024 AnyFeature-VSLAM will be available in the coming weeks. Extensions to RGB-D and Stereo cameras are coming soon, with future work planned for extensions to Visual Inertial systems.

Acknowledgments to: Raul Mur-Artal, Juan D. Tardos, J. M. M. Montiel (ORB-SLAM2), Dorian Galvez-Lopez (DBoW2), Carlos Campos, Richard Elvira and Juan J. Gómez Rodríguez (ORB-SLAM3).

If you want to run AnyFeature-VSLAM on your own data that is not contained in VSLAM-LAB, you can follow these steps.

To ensure all dependencies are properly installed, we suggest using mamba. If you haven't installed mamba yet, please download and set up miniforge, which is a more streamlined installer. This installer will create a "base" environment that includes the conda and mamba package managers.

If you already have a conda installation, you can add mamba by running:

conda install mamba -c conda-forge

Clone the repository:

git clone --recurse-submodules https://github.com/alejandrofontan/AnyFeature-VSLAM.git && cd AnyFeature-VSLAM

Create the environment:

mamba env create -f environment.yml && mamba activate anyfeature

Build AnyFeature-VSLAM:

./build.sh

AnyFeature-VSLAM runs seamlessly in sequences from VSLAM-LAB. If you want to run your own sequences you can format your data as follows. A sequence folder containing 1) an 'rgb' folder with the images, 2) an rgb.txt file with timestamps and paths to the images, and 3) a calibration.yaml file. Try the following toy example:

./bin/mono "sequence_path:docs/toy_sequence"

Additional options:

"Feat:orb32" (orb32/akaze61/brisk48/sift128/anyfeatbin/anyfeatnonbin)
"Vis:1"
"Voc:anyfeature_vocabulary"
"exp_folder:""
"exp_id:0"
"FixRes:0"

python createVocabulary.py --feature_id 0 --rgb_folder ~/BOVISA/2008-09-01/rgb --featSettYamlFile settings/orb32_settings.yaml

AnyFeature-VSLAM builds on ORB-SLAM2 and inherits its release under a GPLv3 license. For a list of all other code/library dependencies (and associated licenses), please see Dependencies.md.

If you found this code/work to be useful in your own research, please considering citing the following:

@inproceedings{fontanRSS2024,
  title={{AnyFeature-VSLAM}: Automating the Usage of Any Chosen Feature into Visual SLAM},
  author={Fontan, Alejandro and Civera, Javier and Milford, Michael},
  booktitle={Robotics: Science and Systems},
  year={2024}
}

[AnyFeature-VSLAM] Alejandro Fontan, Javier Civera and Michael Milford, Automating the Usage of Any Chosen Feature into Visual SLAM, Robotics: Science and Systems, 2024. PDF.

[ORB-SLAM3] Carlos Campos, Richard Elvira, Juan J. Gómez Rodríguez, José M. M. Montiel and Juan D. Tardós, ORB-SLAM3: An Accurate Open-Source Library for Visual, Visual-Inertial and Multi-Map SLAM, IEEE Transactions on Robotics 37(6):1874-1890, Dec. 2021. PDF.

[ORB-SLAM2] Raúl Mur-Artal and Juan D. Tardós. ORB-SLAM2: an Open-Source SLAM System for Monocular, Stereo and RGB-D Cameras. IEEE Transactions on Robotics, vol. 33, no. 5, pp. 1255-1262, 2017. PDF.

[ORB-SLAM] Raúl Mur-Artal, J. M. M. Montiel and Juan D. Tardós. ORB-SLAM: A Versatile and Accurate Monocular SLAM System. IEEE Transactions on Robotics, vol. 31, no. 5, pp. 1147-1163, 2015. (2015 IEEE Transactions on Robotics Best Paper Award). PDF.

[DBoW2 Place Recognizer] Dorian Gálvez-López and Juan D. Tardós. Bags of Binary Words for Fast Place Recognition in Image Sequences. IEEE Transactions on Robotics, vol. 28, no. 5, pp. 1188-1197, 2012. PDF