Unsupervised Learning of 3D Structure from Images
Computer Vision and Pattern Recognition
2018-06-20 v2 Machine Learning
Machine Learning
Abstract
A key goal of computer vision is to recover the underlying 3D structure from 2D observations of the world. In this paper we learn strong deep generative models of 3D structures, and recover these structures from 3D and 2D images via probabilistic inference. We demonstrate high-quality samples and report log-likelihoods on several datasets, including ShapeNet [2], and establish the first benchmarks in the literature. We also show how these models and their inference networks can be trained end-to-end from 2D images. This demonstrates for the first time the feasibility of learning to infer 3D representations of the world in a purely unsupervised manner.
Cite
@article{arxiv.1607.00662,
title = {Unsupervised Learning of 3D Structure from Images},
author = {Danilo Jimenez Rezende and S. M. Ali Eslami and Shakir Mohamed and Peter Battaglia and Max Jaderberg and Nicolas Heess},
journal= {arXiv preprint arXiv:1607.00662},
year = {2018}
}
Comments
Appears in Advances in Neural Information Processing Systems 29 (NIPS 2016)