Evaluating Deep Clustering Algorithms on Non-Categorical 3D CAD Models
This work addresses the underexplored area of deep clustering for 3D shapes, which is important for mechanical engineers and researchers analyzing large 3D shape collections in deep geometric computing.
This paper tackles the problem of evaluating deep clustering algorithms on non-categorical 3D CAD models by creating a benchmark with 252,648 annotated pairwise similarities from 22,968 shapes and proposing an ensemble-based comparison approach to address evaluation challenges.
We introduce the first work on benchmarking and evaluating deep clustering algorithms on large-scale non-categorical 3D CAD models. We first propose a workflow to allow expert mechanical engineers to efficiently annotate 252,648 carefully sampled pairwise CAD model similarities, from a subset of the ABC dataset with 22,968 shapes. Using seven baseline deep clustering methods, we then investigate the fundamental challenges of evaluating clustering methods for non-categorical data. Based on these challenges, we propose a novel and viable ensemble-based clustering comparison approach. This work is the first to directly target the underexplored area of deep clustering algorithms for 3D shapes, and we believe it will be an important building block to analyze and utilize the massive 3D shape collections that are starting to appear in deep geometric computing.