Joint graph entropy knowledge distillation for point cloud classification and robustness against corruptions

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Tian, Zhiqiang, Li, Weigang, Hu, Junwei, Deng, Chunhua
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908561022386176
author Tian, Zhiqiang
Li, Weigang
Hu, Junwei
Deng, Chunhua
author_facet Tian, Zhiqiang
Li, Weigang
Hu, Junwei
Deng, Chunhua
contents Classification tasks in 3D point clouds often assume that class events \replaced{are }{follow }independent and identically distributed (IID), although this assumption destroys the correlation between classes. This \replaced{study }{paper }proposes a classification strategy, \textbf{J}oint \textbf{G}raph \textbf{E}ntropy \textbf{K}nowledge \textbf{D}istillation (JGEKD), suitable for non-independent and identically distributed 3D point cloud data, \replaced{which }{the strategy } achieves knowledge transfer of class correlations through knowledge distillation by constructing a loss function based on joint graph entropy. First\deleted{ly}, we employ joint graphs to capture add{the }hidden relationships between classes\replaced{ and}{,} implement knowledge distillation to train our model by calculating the entropy of add{add }graph.\replaced{ Subsequently}{ Then}, to handle 3D point clouds \deleted{that is }invariant to spatial transformations, we construct \replaced{S}{s}iamese structures and develop two frameworks, self-knowledge distillation and teacher-knowledge distillation, to facilitate information transfer between different transformation forms of the same data. \replaced{In addition}{ Additionally}, we use the above framework to achieve knowledge transfer between point clouds and their corrupted forms, and increase the robustness against corruption of model. Extensive experiments on ScanObject, ModelNet40, ScanntV2\_cls and ModelNet-C demonstrate that the proposed strategy can achieve competitive results.
format Preprint
id arxiv_https___arxiv_org_abs_2509_22150
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Joint graph entropy knowledge distillation for point cloud classification and robustness against corruptions
Tian, Zhiqiang
Li, Weigang
Hu, Junwei
Deng, Chunhua
Computer Vision and Pattern Recognition
Information Retrieval
Classification tasks in 3D point clouds often assume that class events \replaced{are }{follow }independent and identically distributed (IID), although this assumption destroys the correlation between classes. This \replaced{study }{paper }proposes a classification strategy, \textbf{J}oint \textbf{G}raph \textbf{E}ntropy \textbf{K}nowledge \textbf{D}istillation (JGEKD), suitable for non-independent and identically distributed 3D point cloud data, \replaced{which }{the strategy } achieves knowledge transfer of class correlations through knowledge distillation by constructing a loss function based on joint graph entropy. First\deleted{ly}, we employ joint graphs to capture add{the }hidden relationships between classes\replaced{ and}{,} implement knowledge distillation to train our model by calculating the entropy of add{add }graph.\replaced{ Subsequently}{ Then}, to handle 3D point clouds \deleted{that is }invariant to spatial transformations, we construct \replaced{S}{s}iamese structures and develop two frameworks, self-knowledge distillation and teacher-knowledge distillation, to facilitate information transfer between different transformation forms of the same data. \replaced{In addition}{ Additionally}, we use the above framework to achieve knowledge transfer between point clouds and their corrupted forms, and increase the robustness against corruption of model. Extensive experiments on ScanObject, ModelNet40, ScanntV2\_cls and ModelNet-C demonstrate that the proposed strategy can achieve competitive results.
title Joint graph entropy knowledge distillation for point cloud classification and robustness against corruptions
topic Computer Vision and Pattern Recognition
Information Retrieval
url https://arxiv.org/abs/2509.22150