Saved in:
Bibliographic Details
Main Authors: Miandashti, Hanieh Shojaei, Zou, Qianqian, Brenner, Claus
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2411.11935
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909608235237376
author Miandashti, Hanieh Shojaei
Zou, Qianqian
Brenner, Claus
author_facet Miandashti, Hanieh Shojaei
Zou, Qianqian
Brenner, Claus
contents Reliable deep learning models require not only accurate predictions but also well-calibrated confidence estimates to ensure dependable uncertainty estimation. This is crucial in safety-critical applications like autonomous driving, which depend on rapid and precise semantic segmentation of LiDAR point clouds for real-time 3D scene understanding. In this work, we introduce a sampling-free approach for estimating well-calibrated confidence values for classification tasks, achieving alignment with true classification accuracy and significantly reducing inference time compared to sampling-based methods. Our evaluation using the Adaptive Calibration Error (ACE) metric for LiDAR semantic segmentation shows that our approach maintains well-calibrated confidence values while achieving increased processing speed compared to a sampling baseline. Additionally, reliability diagrams reveal that our method produces underconfidence rather than overconfident predictions, an advantage for safety-critical applications. Our sampling-free approach offers well-calibrated and time-efficient predictions for LiDAR scene semantic segmentation.
format Preprint
id arxiv_https___arxiv_org_abs_2411_11935
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Calibrated and Efficient Sampling-Free Confidence Estimation for LiDAR Scene Semantic Segmentation
Miandashti, Hanieh Shojaei
Zou, Qianqian
Brenner, Claus
Computer Vision and Pattern Recognition
Machine Learning
Reliable deep learning models require not only accurate predictions but also well-calibrated confidence estimates to ensure dependable uncertainty estimation. This is crucial in safety-critical applications like autonomous driving, which depend on rapid and precise semantic segmentation of LiDAR point clouds for real-time 3D scene understanding. In this work, we introduce a sampling-free approach for estimating well-calibrated confidence values for classification tasks, achieving alignment with true classification accuracy and significantly reducing inference time compared to sampling-based methods. Our evaluation using the Adaptive Calibration Error (ACE) metric for LiDAR semantic segmentation shows that our approach maintains well-calibrated confidence values while achieving increased processing speed compared to a sampling baseline. Additionally, reliability diagrams reveal that our method produces underconfidence rather than overconfident predictions, an advantage for safety-critical applications. Our sampling-free approach offers well-calibrated and time-efficient predictions for LiDAR scene semantic segmentation.
title Calibrated and Efficient Sampling-Free Confidence Estimation for LiDAR Scene Semantic Segmentation
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2411.11935