EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Enhancing 3D semantic scene completion with refinement module

Dunxing Zhang, Jiachen Lu, Han Yang, Lei Bao, Bo Song

Frontiers in Neurorobotics · 06.03.2026 · 20

DOI: 10.3389/fnbot.2026.1768219 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Frontiers in Neurorobotics, online erschienen am 06.03.2026. Verfasst von Dunxing Zhang und 4 weiteren Autor:innen. Beteiligte Einrichtung: Chair of Robotics, Artificial Intelligence and Real-time Systems, Technical University of Munich, Munich, Germany. Die Arbeit nennt 100 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

We propose ESSC-RM, a plug-and-play Enhancing framework for Semantic Scene Completion with a Refinement Module, which can be seamlessly integrated into existing semantic scene completion (SSC) models. ESSC-RM operates in two phases: a baseline SSC network first produces a coarse voxel prediction, which is subsequently refined by a 3D U-Net–based Prediction Noise-Aware Module (PNAM) and Voxel-level Local Geometry Module (VLGM) under multiscale supervision. Experiments on SemanticKITTI show that ESSC-RM consistently improves semantic prediction performance. When integrated into CGFormer and MonoScene, the mean IoU increases from 16.87 to 17.27% and from 11.08 to 11.51%, respectively. These results demonstrate that ESSC-RM serves as a general refinement framework applicable to a wide range of SSC models. Project page: https://github.com/LuckyMax0722/ESSC-RM and https://github.com/LuckyMax0722/VLGSSC .

Schlagwörter

Stichwörter der Autor:innen

PNAMplug-and-playrefinementsemantic scene completionvison-language guidance

Autor:innen und Institutionen

  • Chair of Robotics, Artificial Intelligence and Real-time Systems, Technical University of Munich, Munich, Germany.
  • National Science Center for Earthquake Engineering, Tianjin University, Tianjin, China.
  • School of Civil Engineering, Tianjin University, Tianjin, China.

Zitieren

APA

Zhang, D., Lu, J., Yang, H., Bao, L., & Song, B. (2026). Enhancing 3D semantic scene completion with refinement module. Frontiers in Neurorobotics, 20. https://doi.org/10.3389/fnbot.2026.1768219

Vancouver

Zhang D, Lu J, Yang H, Bao L, Song B. Enhancing 3D semantic scene completion with refinement module. Frontiers in Neurorobotics. 2026;20. doi:10.3389/fnbot.2026.1768219

BibTeX
@article{Zhang2026,
  author = {Zhang, Dunxing and Lu, Jiachen and Yang, Han and Bao, Lei and Song, Bo},
  title = {Enhancing 3D semantic scene completion with refinement module},
  journal = {Frontiers in Neurorobotics},
  year = {2026},
  volume = {20},
  doi = {10.3389/fnbot.2026.1768219}
}
Kostenlos starten →