Embedded Features for 1D CNN-based Action Recognition on Depth Maps

Jacek Trelinski, Bogdan Kwolek

Abstract

In this paper we present an algorithm for human action recognition using only depth maps. A convolutional autoencoder and Siamese neural network are trained to learn embedded features, encapsulating the content of single depth maps. Afterwards, statistical features and multichannel 1D CNN features are extracted on multivariate time-series of such embedded features to represent actions on depth map sequences. The action recognition is achieved by voting in an ensemble of one-vs-all weak classifiers. We demonstrate experimentally that the proposed algorithm achieves competitive results on UTD-MHAD dataset and outperforms by a large margin the best algorithms on 3D Human-Object Interaction Set (SYSU 3DHOI).

Download


Paper Citation


in Harvard Style

Trelinski J. and Kwolek B. (2021). Embedded Features for 1D CNN-based Action Recognition on Depth Maps.In Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4: VISAPP, ISBN 978-989-758-488-6, pages 536-543. DOI: 10.5220/0010340105360543


in Bibtex Style

@conference{visapp21,
author={Jacek Trelinski and Bogdan Kwolek},
title={Embedded Features for 1D CNN-based Action Recognition on Depth Maps},
booktitle={Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4: VISAPP,},
year={2021},
pages={536-543},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010340105360543},
isbn={978-989-758-488-6},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4: VISAPP,
TI - Embedded Features for 1D CNN-based Action Recognition on Depth Maps
SN - 978-989-758-488-6
AU - Trelinski J.
AU - Kwolek B.
PY - 2021
SP - 536
EP - 543
DO - 10.5220/0010340105360543