3D human pose estimation with siamese equivariant embedding

NEUROCOMPUTING

卷:339

Article

Veges, Marton¹  Varga, Viktor¹  Lorincz, Andras¹ 

[1] Eotvos Lorand Univ, Fac Informat, Budapest, Hungary

关键词: 3D pose estimation;    Siamese network;    Equivariant embedding;   

DOI  :  10.1016/j.neucom.2019.02.029

来源: Elsevier

【 摘 要 】

In monocular 3D human pose estimation a common setup is to first detect 2D positions and then lift the detection into 3D coordinates. Many algorithms suffer from overfitting to camera positions in the training set. We propose a siamese architecture that learns a rotation equivariant hidden representation to reduce the need for data augmentation. Our method is evaluated on multiple databases with different base networks and shows a consistent improvement of error metrics. It achieves state-of-the-art cross-camera error rate among algorithms that use estimated 2D joint coordinates only. (C) 2019 Elsevier B.V. All rights reserved.

【 授权许可】

Free