Pose graph optimization on Cube n=8000, m=22236
4.22Errorchordal+gtsam
Evaluation results
| Method | Links | ||
|---|---|---|---|
| chordal+gtsamInitialization strategy=chordal relaxation2021.03 | 4.22 | 17.9 | |
| gradient+gtsamInitialization strategy=Riemannian gradient descent2021.03 | 4.22 | 26.4 | |
| PyTorch+AutogradInitialization strategy=Riemannian gradient descent2021.03 | 4.22 | 18.3 | |
| OursInitialization strategy=Riemannian gradient descent2021.03 | 4.22 | 1.21 | |
| g2oInitialization strategy=None2021.03 | 5.39 | — | |
| gtsamInitialization strategy=None2021.03 | 6.59 | — |