AI Paper Atlas

Research map / 3D vision / Gaussian splatting

SEGS-SLAM: Structure-enhanced 3D Gaussian Splatting SLAM with Appearance Embedding

Tianci Wen, Zhiang Liu, Yongchun Fang

ICCV 2025 · 3D vision › Gaussian splatting · approach: Compact and efficient Gaussian splatting · 14 citations (OpenAlex)

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
add SEPM (structure-enhanced anchor initialization from ORB-SLAM3 point cloud) (ablation)SEGS-SLAM · TUM RGB-D (RGB-D)PSNR: 21.73 -> 24.20Improved
add SEPM (ablation)SEGS-SLAM · EuRoC MAV (Stereo)PSNR: 11.76 -> 22.91Improved
add SEPM (ablation)SEGS-SLAM · Replica (RGB-D)PSNR: 36.07 -> 38.98Improved
add AfME (pose-conditioned appearance embedding) (ablation)SEGS-SLAM · TUM RGB-D (RGB-D)PSNR: 24.66 -> 26.03Improved
add AfME (ablation)SEGS-SLAM · Replica (RGB-D)PSNR: 39.12 -> 39.42Improved
add FPR (frequency pyramid regularization) (ablation)SEGS-SLAM · TUM RGB-D (Mono)PSNR: 24.91 -> 25.17Improved
replace image-index appearance embedding (AE) with pose-based AfME (ablation)SEGS-SLAM · TUM RGB-D (RGB-D)PSNR: 19.56 -> 26.03Improved
use multi-scale FPR instead of single-scale frequency regularization (ablation)SEGS-SLAM · TUM RGB-D (Mono)PSNR: 23.83 -> 25.17Improved
add low-frequency component to FPR (ablation)SEGS-SLAM · TUM RGB-D (Mono)PSNR: 25.17 -> 24.84Hurt
set FPR scale level to 3 instead of 1, 2 or 4 (ablation)SEGS-SLAM · TUM RGB-D (Mono)PSNR: 23.83 -> 25.17 (level 1 vs 3)Improved
use SEGS-SLAM in place of strongest tracking baselinesSEGS-SLAM · Replica (RGB-D) trackingATE RMSE (cm): 0.430 vs GS-ICP SLAM 0.177No clear effect

Later papers that compared with this method

2 later accepted papers report SEGS-SLAM as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 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