aiqbench takeMachine Learning

This paper introduces PESTO, a novel algorithm for LiDAR point cloud registration that utilizes tetrahedra as environment-agnostic features to handle occlusions effectively. The authors provide rigorous mathematical proofs establishing worst-case bounds on alignment error, distinguishing this work from purely empirical benchmarks. Such formal guarantees are critical for advancing the reliability of autonomous systems and robotics that depend on precise 3D spatial understanding.

Source coveringarXiv

What this isA model's reading of the article and related coverage, written for aiqbench. Not the publisher's words.

PESTO: Formally Correct Registration of LiDAR Point Clouds with Limited Overlap

arXivSeptember 17, 20261 min read

arXiv:2609.18082v1 Announce Type: cross Abstract: In this paper we tackle the problem of aligning LiDAR point clouds also known as the point cloud registration problem. We propose a new algorithm, PESTO, that exploits tetrahedra as "universal features" for LiDAR data, i.e., features that are agnostic to the environment where the LiDAR sensors are deployed. We show empirically that PESTO is competitive with existing solutions for aligning LiDAR point clouds, especially in environments with occlusions. Moreover, we establish PESTO's formal correctness by proving worst-case bounds on the alignment error.

Abstract:In this paper we tackle the problem of aligning LiDAR point clouds also known as the point cloud registration problem. We propose a new algorithm, PESTO, that exploits tetrahedra as "universal features" for LiDAR data, i.e., features that are agnostic to the environment where the LiDAR sensors are deployed. We show empirically that PESTO is competitive with existing solutions for aligning LiDAR point clouds, especially in environments with occlusions. Moreover, we establish PESTO's formal correctness by proving worst-case bounds on the alignment error.

8 pages. Accepted at the 65th IEEE Conference on Decision and Control (CDC), 2026. This version includes appendices with proofs omitted from the conference version

Systems and Control (eess.SY); Robotics (cs.RO)

arXiv:2609.18082 [eess.SY]

(or arXiv:2609.18082v1 [eess.SY] for this version)

https://doi.org/10.48550/arXiv.2609.18082 arXiv-issued DOI via DataCite (pending registration)

Full text under the publisher's open licence · 129 words

Read at arXiv ↗Ask about this article