Indirect deep structured learning for 3D human body shape and pose prediction

Vince Tan, Ignas Budvytis and Roberto Cipolla

Abstract

In this paper we present a novel method for 3D human body shape and pose prediction. Our work is motivated by the need to reduce our reliance on costly-to-obtain ground truth labels. To achieve this, we propose training an encoder-decoder network using a two step procedure as follows. During the first step, a decoder is trained to predict a body silhouette using SMPL [2] (a statistical body shape model) parameters as an input. During the second step, the whole network is trained on real image and corresponding silhouette pairs while the decoder is kept fixed. Such a procedure allows for an indirect learning of body shape and pose parameters from real images without requiring any ground truth parameter data.

Session

Orals - Pose Estimation

Files

Paper (PDF)
Supplementary (PDF)

DOI

10.5244/C.31.15
https://dx.doi.org/10.5244/C.31.15

Citation

Bibtex

            @inproceedings{BMVC2017_15,
                title={Indirect deep structured learning for 3D human body shape and pose prediction},
                author={Vince Tan, Ignas Budvytis and Roberto Cipolla},
                year={2017},
                month={September},
                pages={15.1-15.11},
                articleno={15},
                numpages={11},
                booktitle={Proceedings of the British Machine Vision Conference (BMVC)},
                publisher={BMVA Press},
                editor={Tae-Kyun Kim, Stefanos Zafeiriou, Gabriel Brostow and Krystian Mikolajczyk},
                doi={10.5244/C.31.15},
                isbn={1-901725-60-X},
                url={https://dx.doi.org/10.5244/C.31.15}
            }