Saved in:
Bibliographic Details
Main Authors: You, Jeongbin, Kim, Donggun, Park, Sejun, Oh, Seungsang
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2602.05522
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910012839821312
author You, Jeongbin
Kim, Donggun
Park, Sejun
Oh, Seungsang
author_facet You, Jeongbin
Kim, Donggun
Park, Sejun
Oh, Seungsang
contents Robust 3D point cloud classification is often pursued by scaling up backbones or relying on specialized data augmentation. We instead ask whether structural abstraction alone can improve robustness, and study a simple topology-inspired decomposition based on the Mapper algorithm. We propose Mapper-GIN, a lightweight pipeline that partitions a point cloud into overlapping regions using Mapper (PCA lens, cubical cover, and followed by density-based clustering), constructs a region graph from their overlaps, and performs graph classification with a Graph Isomorphism Network. On the corruption benchmark ModelNet40-C, Mapper-GIN achieves competitive and stable accuracy under Noise and Transformation corruptions with only 0.5M parameters. In contrast to prior approaches that require heavier architectures or additional mechanisms to gain robustness, Mapper-GIN attains strong corruption robustness through simple region-level graph abstraction and GIN message passing. Overall, our results suggest that region-graph structure offers an efficient and interpretable source of robustness for 3D visual recognition.
format Preprint
id arxiv_https___arxiv_org_abs_2602_05522
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Mapper-GIN: Lightweight Structural Graph Abstraction for Corrupted 3D Point Cloud Classification
You, Jeongbin
Kim, Donggun
Park, Sejun
Oh, Seungsang
Computer Vision and Pattern Recognition
Geometric Topology
Robust 3D point cloud classification is often pursued by scaling up backbones or relying on specialized data augmentation. We instead ask whether structural abstraction alone can improve robustness, and study a simple topology-inspired decomposition based on the Mapper algorithm. We propose Mapper-GIN, a lightweight pipeline that partitions a point cloud into overlapping regions using Mapper (PCA lens, cubical cover, and followed by density-based clustering), constructs a region graph from their overlaps, and performs graph classification with a Graph Isomorphism Network. On the corruption benchmark ModelNet40-C, Mapper-GIN achieves competitive and stable accuracy under Noise and Transformation corruptions with only 0.5M parameters. In contrast to prior approaches that require heavier architectures or additional mechanisms to gain robustness, Mapper-GIN attains strong corruption robustness through simple region-level graph abstraction and GIN message passing. Overall, our results suggest that region-graph structure offers an efficient and interpretable source of robustness for 3D visual recognition.
title Mapper-GIN: Lightweight Structural Graph Abstraction for Corrupted 3D Point Cloud Classification
topic Computer Vision and Pattern Recognition
Geometric Topology
url https://arxiv.org/abs/2602.05522