Gaussian Belief Propagation for Bundle Adjustment and SLAM

Python repository for Bundle Adjustment on a Graph Processor at CVPR 2020.

Poplar code for Graphcore's IPU will be released when the Poplar SDK becomes publically available.

Setup

Install dependencies for bundle adjustment visualization:

• rtree (conda install rtree), shapely (conda install shapely), trimesh (pip install trimesh[easy]==2.38.38)

Then, in the project directory, install the gbp module:

pip install -e .

Toy simulations

Estimation of node locations in N-dim space, where nearby nodes are connected by linear factors. Use python ndim_posegraph.py -h for information about optional arguments.

python ndim_posegraph.py

Bundle Adjustment

The bundle adjustment problem is specified with --bal_file. Use python ba.py -h for information about optional arguments.

python ba.py --bal_file data/fr1desk_small.txt

Citation

If you find our work useful in your research, please consider citing:

@InProceedings{OrtizCVPR2020,
author = {Ortiz, Joseph and Pupilli, Mark and Leutenegger, Stefan and Davison, Andrew J.},
title = {Bundle Adjustment on a Graph Processor},
booktitle = {{Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)}},
year = {2020}
}