Saved in:
Bibliographic Details
Main Authors: Kulkarni, Anushka, Dubey, Sarthak
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2604.22065
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915952949460992
author Kulkarni, Anushka
Dubey, Sarthak
author_facet Kulkarni, Anushka
Dubey, Sarthak
contents We present Selective Non-Gaussian Refinement (SNGR), a SLAM framework that augments iSAM2 with targeted nested sampling on windows where Gaussian approximations are likely to fail. We detect such regions using the condition number of joint marginal covariances and selectively refine them using the full nonlinear factor graph likelihood, with a gating mechanism to avoid degradation in multimodal cases. Experiments on range-only SLAM with wrong data association show that SNGR achieves high-precision failure detection and consistent local likelihood improvements while reducing computational cost relative to exhaustive non-Gaussian inference. These results highlight both the promise and the limitations of selective refinement for approximate SLAM posteriors.
format Preprint
id arxiv_https___arxiv_org_abs_2604_22065
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle SNGR: Selective Non-Gaussian Refinement for Ambiguous SLAM Factor Graphs
Kulkarni, Anushka
Dubey, Sarthak
Robotics
Numerical Analysis
We present Selective Non-Gaussian Refinement (SNGR), a SLAM framework that augments iSAM2 with targeted nested sampling on windows where Gaussian approximations are likely to fail. We detect such regions using the condition number of joint marginal covariances and selectively refine them using the full nonlinear factor graph likelihood, with a gating mechanism to avoid degradation in multimodal cases. Experiments on range-only SLAM with wrong data association show that SNGR achieves high-precision failure detection and consistent local likelihood improvements while reducing computational cost relative to exhaustive non-Gaussian inference. These results highlight both the promise and the limitations of selective refinement for approximate SLAM posteriors.
title SNGR: Selective Non-Gaussian Refinement for Ambiguous SLAM Factor Graphs
topic Robotics
Numerical Analysis
url https://arxiv.org/abs/2604.22065