Computer Graphics Laboratory ETH Zurich

ETH

HS-Nets : Estimating Human Body Shape from Silhouettes with Convolutional Neural Networks

E. Dibra, H. Jain, C. Öztireli, R. Ziegler, M. Gross

Proceedings of the Fourth International Conference on 3D Vision, 3DV (Stanford, CA, USA, October 25-28, 2016), pp. 108-117
[Abstract] [BibTeX] [PDF]

Abstract

We represent human body shape estimation from binary silhouettes or shaded images as a regression problem, and describe a novel method to tackle it using CNNs. Utilizing a parametric body model, we train CNNs to learn a global mapping from the input to shape parameters used to reconstruct the shapes of people, in neutral poses, with the application of garment fitting in mind. This results in an accurate, robust and automatic system, orders of magnitude faster than methods we compare to, enabling interactive applications. In addition, we show how to combine silhouettes from two views to improve prediction over a single view. The method is extensively evaluated on thousands of synthetic shapes and real data and compared to stateof-art approaches, clearly outperforming methods based on global fitting and strongly competing with more expensive local fitting based ones.

@inproceedings{DBLP:conf/3dim/DibraJOZG16,
author = {Endri Dibra and
Himanshu Jain and
A. Cengiz {\"{O}}ztireli and
Remo Ziegler and
Markus H. Gross},
title = {HS-Nets: Estimating Human Body Shape from Silhouettes with Convolutional
Neural Networks},
booktitle = {Fourth International Conference on 3D Vision, 3DV 2016, Stanford,
CA, USA, October 25-28, 2016},
pages = {108--117},
year = {2016},
crossref = {DBLP:conf/3dim/2016},
url = {http://dx.doi.org/10.1109/3DV.2016.19},
doi = {10.1109/3DV.2016.19},
timestamp = {Sat, 24 Dec 2016 16:38:47 +0100},
biburl = {http://dblp.uni-trier.de/rec/bib/conf/3dim/DibraJOZG16},
bibsource = {dblp computer science bibliography, http://dblp.org}
}

@proceedings{DBLP:conf/3dim/2016,
title = {Fourth International Conference on 3D Vision, 3DV 2016, Stanford,
CA, USA, October 25-28, 2016},
publisher = {{IEEE} Computer Society},
year = {2016},
url = {http://ieeexplore.ieee.org/xpl/mostRecentIssue.jsp?punumber=7784491},
isbn = {978-1-5090-5407-7},
timestamp = {Fri, 23 Dec 2016 12:26:16 +0100},
biburl = {http://dblp.uni-trier.de/rec/bib/conf/3dim/2016},
bibsource = {dblp computer science bibliography, http://dblp.org}
}

[Download BibTeX]

Downloads

Download Paper
[PDF]