Real-Time LiDAR 3D Semantic Segmentation via Multi-View and Cross-Modal Compact Featuring Two-Branch Knowledge Distillation
Sensors · 15.03.2026 · 26 (6) · 1860
DOI: 10.3390/s26061860 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sensors, online erschienen am 15.03.2026. Verfasst von Yun Zhang und 4 weiteren Autor:innen. Beteiligte Einrichtung: School of Automation, Southeast University, Nanjing 210096, China. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Guangdong Basic and Applied Basic Research Foundation, Taihu Lake Innovation Fund for the School of Future Technology of 385 Southeast University, the Taihu Lake Innovation Fund for the School of Future Technology of Southeast University.
Abstract (Originalsprache)
Simultaneous online mapping and semantic segmentation using handheld scanners supports various environmental inspection and measurement tasks. For such scanners, combing visual and LiDAR data is beneficial for improving the segmentation performance. But the direct fusion of multi-modal and multi-view features faces challenges in terms of both real-time performance and robustness. To address these challenges, this paper proposes a multi-view and cross-modal knowledge distillation method for supporting runtime LiDAR-only semantic segmentation. The proposed method hierarchically compacts multi-view and cross-model priors and distills them into two branches to improve segmentation accuracy. In addition, we design an improved data augmentation technique based on PolarMix for rendering more realistic point cloud scenes. The experimental results on the SemanticKITTI and nuScenes datasets demonstrate that the mIoU of our approach outperforms the state-of-the-art knowledge-distillation-based methods. In addition, mapping experiments using a handheld scanner demonstrate the proposed method’s superior real-time performance and accuracy.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- School of Automation, Southeast University, Nanjing 210096, China.
- Southeast University Shenzhen Research Institute, Shenzhen 518063, China.
- Information and Communication Research Institute, China Electric Power Research Institute Co., Ltd., Nanjing 210000, China.
Förderung
- Guangdong Basic and Applied Basic Research Foundation
- Taihu Lake Innovation Fund for the School of Future Technology of 385 Southeast University
- the Taihu Lake Innovation Fund for the School of Future Technology of Southeast University
Zitieren
APA
Zhang, Y., Qian, K., Zhang, Z., Zhang, M., & Yu, H. (2026). Real-Time LiDAR 3D Semantic Segmentation via Multi-View and Cross-Modal Compact Featuring Two-Branch Knowledge Distillation. Sensors, 26(6), 1860. https://doi.org/10.3390/s26061860
Vancouver
Zhang Y, Qian K, Zhang Z, Zhang M, Yu H. Real-Time LiDAR 3D Semantic Segmentation via Multi-View and Cross-Modal Compact Featuring Two-Branch Knowledge Distillation. Sensors. 2026;26(6):1860. doi:10.3390/s26061860
BibTeX
@article{Zhang2026,
author = {Zhang, Yun and Qian, Kun and Zhang, Zihan and Zhang, Min'ao and Yu, Hai},
title = {Real-Time LiDAR 3D Semantic Segmentation via Multi-View and Cross-Modal Compact Featuring Two-Branch Knowledge Distillation},
journal = {Sensors},
year = {2026},
volume = {26},
number = {6},
pages = {1860},
doi = {10.3390/s26061860}
}
EUVIMED