EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

TGR-T: Truncated-Gaussian-Weighted Reliability for Adaptive Dynamic Thresholding in Weakly Supervised Indoor 3D Point Cloud Segmentation

Ziwei Luo, Xinyue Liu, Jun Jiang, Hanyu Qi, Chen Wang, Zhong Xie, Tao Zeng

ISPRS International Journal of Geo-Information · 04.03.2026 · 15 (3) · 108

DOI: 10.3390/ijgi15030108 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ISPRS International Journal of Geo-Information, online erschienen am 04.03.2026. Verfasst von Ziwei Luo und 6 weiteren Autor:innen. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 68 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Natural Science Foundation of Hubei Province, Sichuan Science and Technology Program, National Natural Science Foundation of China.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Indoor 3D point cloud semantic segmentation is a fundamental task for fine-grained scene understanding and intelligent perception. Due to the prohibitive cost of dense point-wise annotations, weakly supervised learning has emerged as a promising alternative for indoor point cloud segmentation. However, existing weakly supervised methods commonly rely on fixed confidence thresholds for pseudo-label selection, which exhibit limited generalization caused by threshold sensitivity, underutilization of informative low-confidence regions, and progressive noise accumulation during self-training. To address these issues, we propose TGR-T, a weakly supervised framework for indoor 3D point cloud semantic segmentation that incorporates truncated-Gaussian-weighted reliability with adaptive dynamic thresholding. Specifically, a reliability-adaptive dynamic thresholding strategy is introduced to guide pseudo-label selection based on the evolving confidence statistics of unlabeled mini-batches, with exponential moving average smoothing employed to produce stable global estimates and robust separation of reliable and ambiguous regions. To further exploit uncertain regions, a learnable truncated Gaussian weighting function is designed to explicitly model prediction uncertainty within the ambiguous set, providing soft supervision by assigning adaptive weights to low-confidence predictions during optimization. Extensive experimental results demonstrate that the proposed framework significantly enhances the exploitation of unlabeled data under extremely limited supervision: extensive experiments conducted on standard indoor 3D scene benchmarks demonstrate that TGR-T achieves competitive or superior segmentation performance under extremely sparse supervision and can even outperform several fully supervised baselines trained with dense annotations while using only 1% labeled points, thereby substantially narrowing the performance gap between weakly supervised and fully supervised 3D semantic segmentation methods.

Förderung

  • Natural Science Foundation of Hubei Province
  • Sichuan Science and Technology Program
  • National Natural Science Foundation of China
  • Open Fund of Hubei Engineering Research Center for High-Precision Deformation Monitoring with “Beidou + Cloud”
  • Joint Open Fund of the Research Platforms of School of Computer Science, China University of Geosciences, Wuhan

Zitieren

APA

Luo, Z., Liu, X., Jiang, J., Qi, H., Wang, C., Xie, Z., & Zeng, T. (2026). TGR-T: Truncated-Gaussian-Weighted Reliability for Adaptive Dynamic Thresholding in Weakly Supervised Indoor 3D Point Cloud Segmentation. ISPRS International Journal of Geo-Information, 15(3), 108. https://doi.org/10.3390/ijgi15030108

Vancouver

Luo Z, Liu X, Jiang J, Qi H, Wang C, Xie Z, et al. TGR-T: Truncated-Gaussian-Weighted Reliability for Adaptive Dynamic Thresholding in Weakly Supervised Indoor 3D Point Cloud Segmentation. ISPRS International Journal of Geo-Information. 2026;15(3):108. doi:10.3390/ijgi15030108

BibTeX
@article{Luo2026,
  author = {Luo, Ziwei and Liu, Xinyue and Jiang, Jun and Qi, Hanyu and Wang, Chen and Xie, Zhong and Zeng, Tao},
  title = {TGR-T: Truncated-Gaussian-Weighted Reliability for Adaptive Dynamic Thresholding in Weakly Supervised Indoor 3D Point Cloud Segmentation},
  journal = {ISPRS International Journal of Geo-Information},
  year = {2026},
  volume = {15},
  number = {3},
  pages = {108},
  doi = {10.3390/ijgi15030108}
}
Kostenlos starten →