Dockerized LO/LIO algorithms run on the same recordings and scored against ground truth. Every run committed to the results repository appears here, so you can rank algorithms per dataset and see how each one changed from run to run.
Each trajectory is aligned to ground truth with a rigid SE(3) fit, like evo_traj -a, and shown in the XY plane. The view opens on ground truth; click legend entries to add or remove algorithms, and drag to zoom.
The headline value of the selected metric for every run of this dataset. Click legend entries to hide or isolate algorithms; double-click one to show only it.