3D VAE-Attention Network: A Parallel System for Single-view 3D Reconstruction

Hu, Fei; Yang, Xinyan; Zhong, Wei; Ye, Long; Zhang, Qin

3D VAE-Attention Network: A Parallel System for Single-view 3D Reconstruction

Files

053-056.pdf (24.14 MB)

Date

2018

Authors

Hu, Fei
Yang, Xinyan
Zhong, Wei
Ye, Long
Zhang, Qin

Publisher

The Eurographics Association

Abstract

3D object reconstruction from single view image is a challenge task. Due to the fact that the information contained in one isolated image is not sufficient for reasonable 3D shape reconstruction, the existing results on single-view 3D reconstruction always lack marginal voxels. To tackle this problem, we propose a parallel system named 3D VAE-attention network (3VAN) for single view 3D reconstruction. Distinct from the common encoder-decoder structure, the proposed network consists of two parallel branches, 3D-VAE and Attention Network. 3D-VAE completes the general shape reconstruction by an extension of standard VAE model, and Attention Network supplements the missing details by a 3D reconstruction attention network. In the experiments, we verify the feasibility of our 3VAN on the ShapeNet and PASCAL 3D+ datasets. By comparing with the state-of-art methods, the proposed 3VAN can produce more precise 3D object models in terms of both qualitative and quantitative evaluation.

        @inproceedings{10.2312:pg.20181279
,
booktitle = {Pacific Graphics Short Papers
},
editor = {Fu, Hongbo and Ghosh, Abhijeet and Kopf, Johannes
},
title = {{3D VAE-Attention Network: A Parallel System for Single-view 3D Reconstruction
}},
author = {Hu, Fei and 
Yang, Xinyan and 
Zhong, Wei and 
Ye, Long and 
Zhang, Qin
},
year = {2018
},
publisher = {The Eurographics Association
},
ISBN = {978-3-03868-073-4
},
DOI = {10.2312/pg.20181279
}
}