AssemblyNet: A Point Cloud Dataset for Predicting Part Directions in an Exploded Layout
收藏NIAID Data Ecosystem2026-05-01 收录
数据链接：
官方服务：
资源简介：
The AssemblyNet dataset contains 5420 samples of 512-point and 1024-point point clouds of mixed industrial assemblies and LEGO models. The dataset contains classification and regression annotations for predicting which direction a part should be removed in to create an exploded layout. Classification annotations use 26 classes representing 26 world space directions while regression annotations use the actual world space direction as X, Y, and Z values.
创建时间：
2024-03-15
