GEOMetrics: Exploiting Geometric Structure for Graph-Encoded Objects

Edward Smith, Scott Fujimoto, Adriana Romero, David Meger
Proceedings of the 36th International Conference on Machine Learning, PMLR 97:5866-5876, 2019.

Abstract

Mesh models are a promising approach for encoding the structure of 3D objects. Current mesh reconstruction systems predict uniformly distributed vertex locations of a predetermined graph through a series of graph convolutions, leading to compromises with respect to performance or resolution. In this paper, we argue that the graph representation of geometric objects allows for additional structure, which should be leveraged for enhanced reconstruction. Thus, we propose a system which properly benefits from the advantages of the geometric structure of graph-encoded objects by introducing (1) a graph convolutional update preserving vertex information; (2) an adaptive splitting heuristic allowing detail to emerge; and (3) a training objective operating both on the local surfaces defined by vertices as well as the global structure defined by the mesh. Our proposed method is evaluated on the task of 3D object reconstruction from images with the ShapeNet dataset, where we demonstrate state of the art performance, both visually and numerically, while having far smaller space requirements by generating adaptive meshes.

Cite this Paper

BibTeX
@InProceedings{pmlr-v97-smith19a, title = {{GEOM}etrics: Exploiting Geometric Structure for Graph-Encoded Objects}, author = {Smith, Edward and Fujimoto, Scott and Romero, Adriana and Meger, David}, booktitle = {Proceedings of the 36th International Conference on Machine Learning}, pages = {5866--5876}, year = {2019}, editor = {Chaudhuri, Kamalika and Salakhutdinov, Ruslan}, volume = {97}, series = {Proceedings of Machine Learning Research}, month = {09--15 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v97/smith19a/smith19a.pdf}, url = {https://proceedings.mlr.press/v97/smith19a.html}, abstract = {Mesh models are a promising approach for encoding the structure of 3D objects. Current mesh reconstruction systems predict uniformly distributed vertex locations of a predetermined graph through a series of graph convolutions, leading to compromises with respect to performance or resolution. In this paper, we argue that the graph representation of geometric objects allows for additional structure, which should be leveraged for enhanced reconstruction. Thus, we propose a system which properly benefits from the advantages of the geometric structure of graph-encoded objects by introducing (1) a graph convolutional update preserving vertex information; (2) an adaptive splitting heuristic allowing detail to emerge; and (3) a training objective operating both on the local surfaces defined by vertices as well as the global structure defined by the mesh. Our proposed method is evaluated on the task of 3D object reconstruction from images with the ShapeNet dataset, where we demonstrate state of the art performance, both visually and numerically, while having far smaller space requirements by generating adaptive meshes.} }
Endnote
%0 Conference Paper %T GEOMetrics: Exploiting Geometric Structure for Graph-Encoded Objects %A Edward Smith %A Scott Fujimoto %A Adriana Romero %A David Meger %B Proceedings of the 36th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2019 %E Kamalika Chaudhuri %E Ruslan Salakhutdinov %F pmlr-v97-smith19a %I PMLR %P 5866--5876 %U https://proceedings.mlr.press/v97/smith19a.html %V 97 %X Mesh models are a promising approach for encoding the structure of 3D objects. Current mesh reconstruction systems predict uniformly distributed vertex locations of a predetermined graph through a series of graph convolutions, leading to compromises with respect to performance or resolution. In this paper, we argue that the graph representation of geometric objects allows for additional structure, which should be leveraged for enhanced reconstruction. Thus, we propose a system which properly benefits from the advantages of the geometric structure of graph-encoded objects by introducing (1) a graph convolutional update preserving vertex information; (2) an adaptive splitting heuristic allowing detail to emerge; and (3) a training objective operating both on the local surfaces defined by vertices as well as the global structure defined by the mesh. Our proposed method is evaluated on the task of 3D object reconstruction from images with the ShapeNet dataset, where we demonstrate state of the art performance, both visually and numerically, while having far smaller space requirements by generating adaptive meshes.
APA
Smith, E., Fujimoto, S., Romero, A. & Meger, D.. (2019). GEOMetrics: Exploiting Geometric Structure for Graph-Encoded Objects. Proceedings of the 36th International Conference on Machine Learning, in Proceedings of Machine Learning Research 97:5866-5876 Available from https://proceedings.mlr.press/v97/smith19a.html.

Related Material