cs.CVNov 2, 2026

Aerial Content-Aware Sample-Adaptive Classifier for Urban Photogrammetric Point Cloud Segmentation

Authors: Chong Liu, Yonghua Lu, Zhanglin Cheng

Organizations: Southern University of Science and Technology · Shenzhen University of Advanced Technology · Shenzhen VisuCA Key Lab, Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences · Urban Planning & Design Institute of Shenzhen

Abstract

Urban photogrammetric point clouds are commonly processed as local 3D patches, limiting the context available for distinguishing geometrically similar categories. Orthographic projection of the same data yields a co-registered digital orthophoto map (DOM)/RGB-D representation that supports efficient large-context modeling but compresses vertical geometry. Existing projection--point methods mainly fuse these two representations at the feature level while retaining a classifier shared across all samples. We present the Aerial Content-Aware Sample-Adaptive Classifier (ACASAC), which complements feature fusion with decision-level adaptation. A point-aligned aerial region bridges the full aerial crop and local 3D patch. ACASAC models scene-to-local variation within the projection and the representation discrepancy between aligned 2D and 3D supports, then transfers both relations to refine the local point class representation. The refined representation, scene-level aerial context, and fixed fused-head prior jointly form a classifier for the current sample. ACASAC improves the fixed dual-representation head by 2.1 and 2.7 mIoU points on SensatUrban and UrbanBIS, respectively, and by 2.1--2.3 points across four tested encoder combinations.

Figures & tables

Explore similar work

CardsList
  1. COSTA: A Cluster-Centric Paradigm for Annotation-Free Open-Set Semantic Segmentation of Aerial Point Clouds with Domain Shifts

    Aug 19, 2026Yanghong Lin, Li Fang, Tianyu Li +2Graph-Based Label Propagation3D Point Cloud Segmentation

  2. Instance Segmentation and Fine-grained Classification for Urban Buildings with Adaptive Region Dividing and Spatially-Supervised Contrastive Learning

    Sep 17, 2026Weiyuan Zhang, Qi Zhang, Hui HuangContrastive LearningPoint Cloud Classification

  3. Learning Hierarchical Orthogonal Prototypes for Generalized Few-Shot 3D Point Cloud Segmentation

    Mar 20, 2026Yifei Zhao, Fanyu Zhao, Zhongyuan Zhang +33D Point Cloud Segmentation3D Semantic Segmentation