EUVIMEDEUVIMED entdecken
Crossref

A Review of Common Datasets and Algorithms for 3D Gaussian Splatting SLAM in Dynamic Scenes

Advances in Computer, Signals and Systems · 2026 · 10 (1)

DOI: 10.23977/acss.2026.100112 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Advances in Computer, Signals and Systems, 2026.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

: Simultaneous Localization and Mapping (SLAM) is a fundamental technology for autonomous navigation, augmented reality, and virtual reality. However, dynamic factors widely present in real-world environments, such as pedestrians, vehicles, and movable objects, severely violate the static-world assumption adopted by traditional SLAM methods, posing significant challenges to localization accuracy, map consistency, and long-term stability. In recent years, 3D Gaussian Splatting (3DGS) has attracted increasing attention due to its compact and efficient scene representation, favorable differentiability, and excellent real-time rendering performance. It provides a novel technical paradigm for SLAM in dynamic environments and has gradually become a research hotspot. This paper presents a systematic review of 3D Gaussian SLAM algorithms in dynamic scenes. First, commonly used benchmark datasets are analyzed, including the TUM and BONN datasets for indoor dynamic environments and the KITTI dataset for outdoor scenarios. Comparisons are conducted in terms of scene scale, dynamic object types, and evaluation protocols. Second, commonly used evaluation metrics for pose accuracy, rendering quality, and system efficiency are summarized. Third, representative dynamic-scene 3D Gaussian SLAM algorithms are reviewed, and their core ideas and technical characteristics for handling dynamic interference are systematically analyzed. Finally, existing challenges are discussed, and future research directions are outlined.

Zitieren

APA

(2026). A Review of Common Datasets and Algorithms for 3D Gaussian Splatting SLAM in Dynamic Scenes. Advances in Computer, Signals and Systems, 10(1). https://doi.org/10.23977/acss.2026.100112

Vancouver

A Review of Common Datasets and Algorithms for 3D Gaussian Splatting SLAM in Dynamic Scenes. Advances in Computer, Signals and Systems. 2026;10(1). doi:10.23977/acss.2026.100112

BibTeX
@article{euvimed2026,
  title = {A Review of Common Datasets and Algorithms for 3D Gaussian Splatting SLAM in Dynamic Scenes},
  journal = {Advances in Computer, Signals and Systems},
  year = {2026},
  volume = {10},
  number = {1},
  doi = {10.23977/acss.2026.100112}
}
Kostenlos starten →