Multiresolution Tree Networks for 3D Point Cloud Processing
About
We present multiresolution tree-structured networks to process point clouds for 3D shape understanding and generation tasks. Our network represents a 3D shape as a set of locality-preserving 1D ordered list of points at multiple resolutions. This allows efficient feed-forward processing through 1D convolutions, coarse-to-fine analysis through a multi-grid architecture, and it leads to faster convergence and small memory footprint during training. The proposed tree-structured encoders can be used to classify shapes and outperform existing point-based architectures on shape classification benchmarks, while tree-structured decoders can be used for generating point clouds directly and they outperform existing approaches for image-to-shape inference tasks learned using the ShapeNet dataset. Our model also allows unsupervised learning of point-cloud based shapes by using a variational autoencoder, leading to higher-quality generated shapes.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| 3D Point Cloud Classification | ModelNet40 (test) | OA91.2 | 297 | |
| 3D Shape Classification | ModelNet40 (test) | Accuracy91.2 | 227 | |
| Object Classification | ModelNet40 (test) | -- | 180 | |
| Classification | ModelNet40 (test) | Accuracy86.4 | 99 | |
| Unsupervised Representation Learning | ModelNet40 (test) | Accuracy86.4 | 13 |