Towards Explaining Uncertainty Estimates in Point Cloud Registration

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Qin, Ziyuan, Lee, Jongseok, Triebel, Rudolph
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910766403158016
author Qin, Ziyuan
Lee, Jongseok
Triebel, Rudolph
author_facet Qin, Ziyuan
Lee, Jongseok
Triebel, Rudolph
contents Iterative Closest Point (ICP) is a commonly used algorithm to estimate transformation between two point clouds. The key idea of this work is to leverage recent advances in explainable AI for probabilistic ICP methods that provide uncertainty estimates. Concretely, we propose a method that can explain why a probabilistic ICP method produced a particular output. Our method is based on kernel SHAP (SHapley Additive exPlanations). With this, we assign an importance value to common sources of uncertainty in ICP such as sensor noise, occlusion, and ambiguous environments. The results of the experiment show that this explanation method can reasonably explain the uncertainty sources, providing a step towards robots that know when and why they failed in a human interpretable manner
format Preprint
id arxiv_https___arxiv_org_abs_2412_20612
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Towards Explaining Uncertainty Estimates in Point Cloud Registration
Qin, Ziyuan
Lee, Jongseok
Triebel, Rudolph
Robotics
Artificial Intelligence
Iterative Closest Point (ICP) is a commonly used algorithm to estimate transformation between two point clouds. The key idea of this work is to leverage recent advances in explainable AI for probabilistic ICP methods that provide uncertainty estimates. Concretely, we propose a method that can explain why a probabilistic ICP method produced a particular output. Our method is based on kernel SHAP (SHapley Additive exPlanations). With this, we assign an importance value to common sources of uncertainty in ICP such as sensor noise, occlusion, and ambiguous environments. The results of the experiment show that this explanation method can reasonably explain the uncertainty sources, providing a step towards robots that know when and why they failed in a human interpretable manner
title Towards Explaining Uncertainty Estimates in Point Cloud Registration
topic Robotics
Artificial Intelligence
url https://arxiv.org/abs/2412.20612