Learning Representations and Generative Models for 3D Point Clouds

Panos Achlioptas, Olga Diamanti, Ioannis Mitliagkas, Leonidas Guibas
Proceedings of the 35th International Conference on Machine Learning, PMLR 80:40-49, 2018.

Abstract

Three-dimensional geometric data offer an excellent domain for studying representation learning and generative modeling. In this paper, we look at geometric data represented as point clouds. We introduce a deep AutoEncoder (AE) network with state-of-the-art reconstruction quality and generalization ability. The learned representations outperform existing methods on 3D recognition tasks and enable shape editing via simple algebraic manipulations, such as semantic part editing, shape analogies and shape interpolation, as well as shape completion. We perform a thorough study of different generative models including GANs operating on the raw point clouds, significantly improved GANs trained in the fixed latent space of our AEs, and Gaussian Mixture Models (GMMs). To quantitatively evaluate generative models we introduce measures of sample fidelity and diversity based on matchings between sets of point clouds. Interestingly, our evaluation of generalization, fidelity and diversity reveals that GMMs trained in the latent space of our AEs yield the best results overall.

Cite this Paper

BibTeX
@InProceedings{pmlr-v80-achlioptas18a, title = {Learning Representations and Generative Models for 3{D} Point Clouds}, author = {Achlioptas, Panos and Diamanti, Olga and Mitliagkas, Ioannis and Guibas, Leonidas}, booktitle = {Proceedings of the 35th International Conference on Machine Learning}, pages = {40--49}, year = {2018}, editor = {Dy, Jennifer and Krause, Andreas}, volume = {80}, series = {Proceedings of Machine Learning Research}, month = {10--15 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v80/achlioptas18a/achlioptas18a.pdf}, url = {https://proceedings.mlr.press/v80/achlioptas18a.html}, abstract = {Three-dimensional geometric data offer an excellent domain for studying representation learning and generative modeling. In this paper, we look at geometric data represented as point clouds. We introduce a deep AutoEncoder (AE) network with state-of-the-art reconstruction quality and generalization ability. The learned representations outperform existing methods on 3D recognition tasks and enable shape editing via simple algebraic manipulations, such as semantic part editing, shape analogies and shape interpolation, as well as shape completion. We perform a thorough study of different generative models including GANs operating on the raw point clouds, significantly improved GANs trained in the fixed latent space of our AEs, and Gaussian Mixture Models (GMMs). To quantitatively evaluate generative models we introduce measures of sample fidelity and diversity based on matchings between sets of point clouds. Interestingly, our evaluation of generalization, fidelity and diversity reveals that GMMs trained in the latent space of our AEs yield the best results overall.} }
Endnote
%0 Conference Paper %T Learning Representations and Generative Models for 3D Point Clouds %A Panos Achlioptas %A Olga Diamanti %A Ioannis Mitliagkas %A Leonidas Guibas %B Proceedings of the 35th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2018 %E Jennifer Dy %E Andreas Krause %F pmlr-v80-achlioptas18a %I PMLR %P 40--49 %U https://proceedings.mlr.press/v80/achlioptas18a.html %V 80 %X Three-dimensional geometric data offer an excellent domain for studying representation learning and generative modeling. In this paper, we look at geometric data represented as point clouds. We introduce a deep AutoEncoder (AE) network with state-of-the-art reconstruction quality and generalization ability. The learned representations outperform existing methods on 3D recognition tasks and enable shape editing via simple algebraic manipulations, such as semantic part editing, shape analogies and shape interpolation, as well as shape completion. We perform a thorough study of different generative models including GANs operating on the raw point clouds, significantly improved GANs trained in the fixed latent space of our AEs, and Gaussian Mixture Models (GMMs). To quantitatively evaluate generative models we introduce measures of sample fidelity and diversity based on matchings between sets of point clouds. Interestingly, our evaluation of generalization, fidelity and diversity reveals that GMMs trained in the latent space of our AEs yield the best results overall.
APA
Achlioptas, P., Diamanti, O., Mitliagkas, I. & Guibas, L.. (2018). Learning Representations and Generative Models for 3D Point Clouds. Proceedings of the 35th International Conference on Machine Learning, in Proceedings of Machine Learning Research 80:40-49 Available from https://proceedings.mlr.press/v80/achlioptas18a.html.

Related Material