Different from 2D images that have a dominant representation as pixel arrays, 3D data has multiple popular representations, such as point cloud, mesh, volumetric field, multi-view images and parametric models. So far, I feel parametric models get the least attention in ML. I am not sure if this is true for both academic group and industrial group.
A summary from timzhang642(github) ** very good! Thanks!
Stanford CS468 -m achine learning for 3D data
Using machine learning to extract 3D data ** A forum post with interesting idea and several cool youtube videos attached by another forum users, sticky-figure actions…
PDF - machine_learning_techniques_for_geometric_modeling
PDF - Deep Learning for Advanced 3D Printing
PDF - 3D ShapeNets for 2.5D Object Recognition
网友评论