Published on Fri Jun 07 2019

Latent feature disentanglement for 3D meshes

Jake Levinson, Avneesh Sud, Ameesh Makadia

Generative modeling of 3D shapes has become an important problem due to its relevance to Computer Vision, Graphics, and VR. We introduce a supervised generative 3D mesh model that disentangles the latent shape representation into independent generative factors.

0
0
0
Abstract

Generative modeling of 3D shapes has become an important problem due to its relevance to many applications across Computer Vision, Graphics, and VR. In this paper we build upon recently introduced 3D mesh-convolutional Variational AutoEncoders which have shown great promise for learning rich representations of deformable 3D shapes. We introduce a supervised generative 3D mesh model that disentangles the latent shape representation into independent generative factors. Our extensive experimental analysis shows that learning an explicitly disentangled representation can both improve random shape generation as well as successfully address downstream tasks such as pose and shape transfer, shape-invariant temporal synchronization, and pose-invariant shape matching.

Sat Feb 27 2021
Machine Learning
Disentangling Geometric Deformation Spaces in Generative Latent Shape Models
A complete representation of 3D objects requires characterizing the space of deformations in an interpretable manner. In this work, we improve on aprior generative model of geometric disentanglement for 3D shapes.
0
0
0
Sun Aug 18 2019
Machine Learning
Geometric Disentanglement for Generative Latent Shape Models
Representing 3D shape is a fundamental problem in artificial intelligence. One avenue has recently begun to be explored is the use of latent representations of generative models. We show that the resulting representation exhibits intuitive and interpretable behavior.
0
0
0
Fri Aug 06 2021
Computer Vision
GLASS: Geometric Latent Augmentation for Shape Spaces
We use geometrically motivated energies to augment and thus boost a sparse collection of example (training) models. Our framework allows us to train generative 3Dmodels even with a small set of good quality 3D models, which are typically hard to curate.
3
0
1
Mon Nov 16 2020
Machine Learning
Cycle-Consistent Generative Rendering for 2D-3D Modality Translation
For humans, visual understanding is inherently generative: given a 3D shape, we can postulate how it would look in the world. In the context of computer vision, this corresponds to a learnable module. We learn such a module while being conscious of the difficulties in obtaining large paired 2D-3D datasets.
0
0
0
Fri Dec 01 2017
Computer Vision
Deformable Shape Completion with Graph Convolutional Autoencoders
The problem of reconstructing a 3D object from a single or multiple partial scans has received increasing attention in recent years. Unlike the majority of existing approaches, our method focuses on objects that can undergo non-rigid deformations.
0
0
0
Wed May 08 2019
Artificial Intelligence
Neural 3D Morphable Models: Spiral Convolutional Networks for 3D Shape Representation Learning and Generation
Generative models for 3D geometric data arise in many important applications. In this work, we introduce a novel graph convolutional operator, acting directly on the 3D mesh. We demonstrate state-of-the-art results on a variety of 3D shape datasets.
0
0
0