Pith. sign in

REVIEW 3 cited by

SGS-SLAM: Semantic Gaussian Splatting For Neural Dense SLAM

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2402.03246 v6 pith:AZCF42T6 submitted 2024-02-05 cs.CV cs.AIcs.RO

Mingrui Li , Shuhong Liu , Heng Zhou , Guohao Zhu , Na Cheng , Tianchen Deng , Hongyu Wang This is my paper
classification cs.CVcs.AIcs.RO
keywords semanticsgs-slamslamgaussiangeometryneuralobject-leveloptimization
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

We present SGS-SLAM, the first semantic visual SLAM system based on Gaussian Splatting. It incorporates appearance, geometry, and semantic features through multi-channel optimization, addressing the oversmoothing limitations of neural implicit SLAM systems in high-quality rendering, scene understanding, and object-level geometry. We introduce a unique semantic feature loss that effectively compensates for the shortcomings of traditional depth and color losses in object optimization. Through a semantic-guided keyframe selection strategy, we prevent erroneous reconstructions caused by cumulative errors. Extensive experiments demonstrate that SGS-SLAM delivers state-of-the-art performance in camera pose estimation, map reconstruction, precise semantic segmentation, and object-level geometric accuracy, while ensuring real-time rendering capabilities.

Discussion (0). Sign in to comment.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. VTGaussian-SLAM: RGBD SLAM for Large Scale Scenes with Splatting View-Tied 3D Gaussians

    cs.CV 2025-06 conditional novelty 6.0 of 10

    A new RGBD SLAM representation ties Gaussian positions to depth pixels, leaving only color, radius, and opacity learnable, enabling local-only optimization and higher rendering quality on several benchmarks.

  2. LRSLAM: Low-rank Representation of Signed Distance Fields in Dense Visual SLAM System

    cs.CV 2025-06 conditional novelty 5.0 of 10

    LRSLAM couples CP decomposition for geometry with a new Six-axis decomposition for appearance, yielding a dense RGB-D SLAM with linear memory growth and accuracy competitive with ESLAM.

  3. LEG-SLAM: Real-Time Language-Enhanced Gaussian Splatting for SLAM

    cs.CV 2025-06 conditional novelty 5.0 of 10

    LEG-SLAM is a real-time RGB-D SLAM that jointly renders photorealistic images and open-vocabulary semantic masks by distilling PCA-compressed DINOv2 features into 3D Gaussians.

Pith tools