Deep Modeling of Non-Gaussian Aleatoric Uncertainty

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Acharya, Aastha, Lee, Caleb, D'Alonzo, Marissa, Shamwell, Jared, Ahmed, Nisar R., Russell, Rebecca
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915174340886528
author Acharya, Aastha
Lee, Caleb
D'Alonzo, Marissa
Shamwell, Jared
Ahmed, Nisar R.
Russell, Rebecca
author_facet Acharya, Aastha
Lee, Caleb
D'Alonzo, Marissa
Shamwell, Jared
Ahmed, Nisar R.
Russell, Rebecca
contents Deep learning offers promising new ways to accurately model aleatoric uncertainty in robotic state estimation systems, particularly when the uncertainty distributions do not conform to traditional assumptions of being fixed and Gaussian. In this study, we formulate and evaluate three fundamental deep learning approaches for conditional probability density modeling to quantify non-Gaussian aleatoric uncertainty: parametric, discretized, and generative modeling. We systematically compare the respective strengths and weaknesses of these three methods on simulated non-Gaussian densities as well as on real-world terrain-relative navigation data. Our results show that these deep learning methods can accurately capture complex uncertainty patterns, highlighting their potential for improving the reliability and robustness of estimation systems.
format Preprint
id arxiv_https___arxiv_org_abs_2405_20513
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Deep Modeling of Non-Gaussian Aleatoric Uncertainty
Acharya, Aastha
Lee, Caleb
D'Alonzo, Marissa
Shamwell, Jared
Ahmed, Nisar R.
Russell, Rebecca
Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
Robotics
Deep learning offers promising new ways to accurately model aleatoric uncertainty in robotic state estimation systems, particularly when the uncertainty distributions do not conform to traditional assumptions of being fixed and Gaussian. In this study, we formulate and evaluate three fundamental deep learning approaches for conditional probability density modeling to quantify non-Gaussian aleatoric uncertainty: parametric, discretized, and generative modeling. We systematically compare the respective strengths and weaknesses of these three methods on simulated non-Gaussian densities as well as on real-world terrain-relative navigation data. Our results show that these deep learning methods can accurately capture complex uncertainty patterns, highlighting their potential for improving the reliability and robustness of estimation systems.
title Deep Modeling of Non-Gaussian Aleatoric Uncertainty
topic Machine Learning
Artificial Intelligence
Computer Vision and Pattern Recognition
Robotics
url https://arxiv.org/abs/2405.20513