Abstract:We developed graph analysis tools processing on Maximal Disjoint Ball Decomposition (MDBD) for shape analyzing tasks, and the consistency of results across any valid geometric representations is guaranteed by the nature of MDBD.We designed a DNN based on graph CNN that achieves a comparable classification accuracy with that of other leading classifiers on a widely used benchmark dataset, but with a significantly lower number of parameters.We proposed a graph kernel based on random walks that exploits the hierarchical structure of MDBD and has linear time complexity.We implemented a graph clustering technique (Adversarially Regularized Variational Graph Autoencoder + KMeans), and showed that we can segment geometric models within an unsupervised machine learning framework.By combining the graph kernel and the model segmentation technique, we established a substructure matching method which leads to novel, powerful, and representation-agnostic feature recognition algorithms that can hardly be built with existing shape descriptors.Practically all existing shape analysis and processing algorithms have been developed for specific geometric representations of 3D models. However, the product development process always involves a large number of often incompatible geometric representations tailored to specific computational tasks that take place during this process. Consequently, a substantial effort has been expended to develop robust geometric data translation and conversion algorithms, but the existing methods have well known limitations.The Maximal Disjoint Ball Decomposition (MDBD) was recently defined as a unique and stable geometric construction and used to define universal shape descriptors based on MDBD's associated contact graph. In this paper, we demonstrate that by applying graph analysis tools to MDBD in conjunction with graph convolutional neural networks and graph kernels, one can effectively develop methods to perform similarity, segmentation and solid feature recognition from geometric models regardless of their native geometric representation. We show that our representation-agnostic approach achieves comparable performance with state-of-the-art geometric processing methods on standard yet heterogeneous benchmark datasets while supporting all valid geometric representations.Download : Download high-res image (224KB)Download : Download full-size image

Graph-Based Shape Analysis for Heterogeneous Geometric Datasets: Similarity, Retrieval and Substructure Matching

Shape Representation and Recognition in High Dimensional Feature Space

Smart Computing for Large Scale Visual Data Sensing and Processing

Learning Point Cloud Shapes with Geometric and Topological Structures.

A Unified Feature Representation and Learning Framework for 3D Shape

Shape Decomposition for Graph Representation

Non-rigid 3D Shape Retrieval Based on Multi-Scale Graphical Image and Joint Bayesian

3D Shape Knowledge Graph for Cross‐domain 3D Shape Retrieval

Manifold embedding for shape analysis

Data‐Driven Shape Analysis and Processing

Exploring Generalized Shape Analysis by Topological Representations

The Generalized Shape Distributions for Shape Matching and Analysis

Shape Retrieval And Classification Based On Geodesic Paths In Skeleton Graphs

Representations, Metrics and Statistics For Shape Analysis of Elastic Graphs

Multi-graph Convolutional Network for Unsupervised 3D Shape Retrieval

Shape Classification with a Vertex Clustering Graph Kernel

3D Shape Recognition and Retrieval Based on Multi-Modality Deep Learning

Discriminative Sketch-Based 3d Model Retrieval Via Robust Shape Matching

Shape Similarity Assessment Based On Partial Feature Aggregation And Ranking Lists

A Shape Recognition Method Based on Graph- and Line-Contexts

Hierarchical Graphical Models for Multigroup Shape Analysis using Expectation Maximization with Sampling in Kendall's Shape Space