AI Paper Atlas

Research map / 3D vision / Gaussian splatting

AERGS-SLAM: Auto-Exposure-Robust Stereo 3D Gaussian Splatting SLAM

Zhiyu Zhou, Feng Hui, Yu Liu

CVPR 2026 · 3D vision › Gaussian splatting · approach: Regularization and primitive design for 3DGS

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace ORB features with learning-based illumination-robust features (IRL) (ablation)Photo-SLAM based pipeline · EuRoC (average)ATE RMSE: 0.164 -> 0.072Improved
replace ORB features with learning-based illumination-robust features (IRL) (ablation)Photo-SLAM based pipeline · Self-collected (average)ATE RMSE: 1.754 -> 0.199Improved
add IRL (better poses and initialization) (ablation)Photo-SLAM based pipeline · EuRoC (average)PSNR: 11.76 -> 14.76Improved
add camera exposure network (CEN) (ablation)AERGS-SLAM · EuRoC (average)PSNR: 15.10 -> 21.11Improved
add camera exposure network (CEN) (ablation)AERGS-SLAM · Self-collected (average)PSNR: 19.69 -> 20.06Improved
add time-aware coarse-to-fine optimization (CTFO) (ablation)AERGS-SLAM · EuRoC (average)PSNR: 20.48 -> 21.11Improved
add time-aware coarse-to-fine optimization (CTFO) (ablation)AERGS-SLAM · Self-collected (average)PSNR: 20.05 -> 20.06No clear effect
use CEN instead of HDR-GS exposure modeling (ablation)AERGS-SLAM · EuRoC V102 (brightness-perturbed)PSNR: 20.31 -> 23.55Improved
use CEN instead of HDR-GS exposure modeling (ablation)AERGS-SLAM · Self-collected S2PSNR: 21.93 -> 21.60Hurt
map per-image radiance map via CEN instead of per-Gaussian radiance (HDR-GS) (ablation)AERGS-SLAM · Exposure-controlled renderingrendering speed (FPS): 416 -> 3700Improved
use learning-based features instead of ORB in decoupled 3DGS SLAMPhoto-SLAM · EuRoC MAV (brightness-perturbed)run successImproved
decoupled localization and mapping instead of coupled (MonoGS)3DGS SLAM · EuRoC MAV MH01 (brightness-perturbed)ATE RMSE: 0.089 -> 0.029Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy