SVN-ICP : Uncertainty Estimation of ICP-based LiDAR Odometry using Stein Variational Newton
Shiping Ma, Haoming Zhang, Marc Toussaint · 2025
Read SVN-ICP : Uncertainty Estimation of ICP-based LiDAR Odometry using Stein Variational Newton by Shiping Ma, Haoming Zhang, Marc Toussaint (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-08069.txt.
Description
Shiping Ma and 2 co-authors add honest doubt to laser-based robot tracking in a 2025 robotics paper. The standard method aligns one laser scan with the next to track motion. It gives a pose, but no sense of how wrong that pose may be. Fusion systems need that doubt. Bare halls and tunnels make laser data weak, and doubt grows there.
Their method is called SVN-ICP. It treats the pose as a spread of options, not a single point. A set of samples stands in for the full spread. Noise levels are learned from the data, not set by hand. Tests plug the method into a Kalman filter, the standard tool for merging sensors. The aim is a pose that other sensors can trust at the right weight.
Catalog record
| CREATOR | Shiping Ma, Haoming Zhang, Marc Toussaint |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 46 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-08069