TR2017-193
FoldingNet: Interpretable Unsupervised Learning on 3D Point Clouds
-
- "FoldingNet: Interpretable Unsupervised Learning on 3D Point Clouds", arXiv, December 2017. ,
-
Research Areas:
Abstract:
Recent deep networks that directly handle points in a point set, e.g., PointNet, have been state-of-the-art for supervised semantic learning tasks on point clouds such as classification and segmentation. In this work, a novel endto-end deep auto-encoder is proposed to address unsupervised learning challenges on point clouds. On the encoder side, a graph-based enhancement is enforced to promote local structures on top of PointNet. Then, a novel folding based approach is proposed in the decoder, which folds a 2D grid onto the underlying 3D object surface of a point cloud. The proposed decoder only uses about 7% parameters of a decoder with fully-connected neural networks, yet leads to a more discriminative representation that achieves higher linear SVM classification accuracy than the benchmark. In addition, the proposed decoder structure is shown, in theory, to be a generic architecture that is able to reconstruct an arbitrary point cloud from a 2D grid. Finally, this folding-based decoder is interpretable since the reconstruction could be viewed as a fine granular warping from the 2D grid to the point cloud surface.
Software & Data Downloads
Related Publication
- @inproceedings{Yang2018jun,
- author = {Yang, Yaoqing and Feng, Chen and Shen, Yiru and Tian, Dong},
- title = {FoldingNet: Point Cloud Auto-encoder via Deep Grid Deformation},
- booktitle = {IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
- year = 2018,
- month = jun,
- doi = {10.1109/CVPR.2018.00029},
- url = {https://www.merl.com/publications/TR2018-042}
- }