Research of Comparison on Convolution Network and BP Network Based on Human Body Attitude Recognition

DOI

10.2991/jimec-18.2018.36How to use a DOI?

Keywords

Acceleration Sensor; Human motion recognition; BP Network; Convolution Network; Python; TensorFlow; Deep Learning

Abstract

In this thesis, after collecting data for human wearable acceleration sensor, human posture is recognized by using traditional BP network and convolution network based on Tensorflow in Python. Through respectively introducing and using BP network and convolution network, we then made a comparison experiment. It had become clear that the effect of using convolution network is much better than using BP network for the posture mentioned in the article. Its accuracy rate is 75%.

Copyright

© 2019, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Yang Du
AU  - Jiaxin Tian
AU  - Binghong Zhan
AU  - Fei Guo
PY  - 2018/12
DA  - 2018/12
TI  - Research of Comparison on Convolution Network and BP Network Based on Human Body Attitude Recognition
BT  - Proceedings of the 2018 3rd Joint International Information Technology，Mechanical and Electronic Engineering Conference (JIMEC 2018)
PB  - Atlantis Press
SP  - 169
EP  - 172
SN  - 2589-4943
UR  - https://doi.org/10.2991/jimec-18.2018.36
DO  - 10.2991/jimec-18.2018.36
ID  - Du2018/12
ER  -