The Effect of Axis-Wise Triaxial Acceleration Data Fusion in CNN-Based Human Activity Recognition

Xinxin HAN; Jian YE; Jia LUO; Haiying ZHOU

doi:10.1587/transinf.2018EDP7409

The Effect of Axis-Wise Triaxial Acceleration Data Fusion in CNN-Based Human Activity Recognition

Xinxin HAN, Jian YE, Jia LUO, Haiying ZHOU

Full Text Views

0

Cite this

Summary :

The triaxial accelerometer is one of the most important sensors for human activity recognition (HAR). It has been observed that the relations between the axes of a triaxial accelerometer plays a significant role in improving the accuracy of activity recognition. However, the existing research rarely focuses on these relations, but rather on the fusion of multiple sensors. In this paper, we propose a data fusion-based convolutional neural network (CNN) approach to effectively use the relations between the axes. We design a single-channel data fusion method and multichannel data fusion method in consideration of the diversified formats of sensor data. After obtaining the fused data, a CNN is used to extract the features and perform classification. The experiments show that the proposed approach has an advantage over the CNN in accuracy. Moreover, the single-channel model achieves an accuracy of 98.83% with the WISDM dataset, which is higher than that of state-of-the-art methods.

Publication: IEICE TRANSACTIONS on Information Vol.E103-D No.4 pp.813-824

Publication Date: 2020/04/01

Publicized: 2020/01/14

Online ISSN: 1745-1361

DOI: 10.1587/transinf.2018EDP7409

Type of Manuscript: PAPER

Category: Artificial Intelligence, Data Mining

Authors

Xinxin HAN
  North University of China,Chinese Academy of Sciences
Jian YE
  Chinese Academy of Sciences,Beijing Key Laboratory of Mobile Computing and Pervasive Device
Jia LUO
  Beijing University of Technology
Haiying ZHOU
  North University of China

Keyword

triaxial accelerometer, human activity recognition, data fusion, convolutional neural network

Cite this

Copy

Xinxin HAN, Jian YE, Jia LUO, Haiying ZHOU, "The Effect of Axis-Wise Triaxial Acceleration Data Fusion in CNN-Based Human Activity Recognition" in IEICE TRANSACTIONS on Information, vol. E103-D, no. 4, pp. 813-824, April 2020, doi: 10.1587/transinf.2018EDP7409.
Abstract: The triaxial accelerometer is one of the most important sensors for human activity recognition (HAR). It has been observed that the relations between the axes of a triaxial accelerometer plays a significant role in improving the accuracy of activity recognition. However, the existing research rarely focuses on these relations, but rather on the fusion of multiple sensors. In this paper, we propose a data fusion-based convolutional neural network (CNN) approach to effectively use the relations between the axes. We design a single-channel data fusion method and multichannel data fusion method in consideration of the diversified formats of sensor data. After obtaining the fused data, a CNN is used to extract the features and perform classification. The experiments show that the proposed approach has an advantage over the CNN in accuracy. Moreover, the single-channel model achieves an accuracy of 98.83% with the WISDM dataset, which is higher than that of state-of-the-art methods.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2018EDP7409/_p

Copy

@ARTICLE{e103-d_4_813,
author={Xinxin HAN, Jian YE, Jia LUO, Haiying ZHOU, },
journal={IEICE TRANSACTIONS on Information},
title={The Effect of Axis-Wise Triaxial Acceleration Data Fusion in CNN-Based Human Activity Recognition},
year={2020},
volume={E103-D},
number={4},
pages={813-824},
abstract={The triaxial accelerometer is one of the most important sensors for human activity recognition (HAR). It has been observed that the relations between the axes of a triaxial accelerometer plays a significant role in improving the accuracy of activity recognition. However, the existing research rarely focuses on these relations, but rather on the fusion of multiple sensors. In this paper, we propose a data fusion-based convolutional neural network (CNN) approach to effectively use the relations between the axes. We design a single-channel data fusion method and multichannel data fusion method in consideration of the diversified formats of sensor data. After obtaining the fused data, a CNN is used to extract the features and perform classification. The experiments show that the proposed approach has an advantage over the CNN in accuracy. Moreover, the single-channel model achieves an accuracy of 98.83% with the WISDM dataset, which is higher than that of state-of-the-art methods.},
keywords={},
doi={10.1587/transinf.2018EDP7409},
ISSN={1745-1361},
month={April},}

Copy

TY - JOUR
TI - The Effect of Axis-Wise Triaxial Acceleration Data Fusion in CNN-Based Human Activity Recognition
T2 - IEICE TRANSACTIONS on Information
SP - 813
EP - 824
AU - Xinxin HAN
AU - Jian YE
AU - Jia LUO
AU - Haiying ZHOU
PY - 2020
DO - 10.1587/transinf.2018EDP7409
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E103-D
IS - 4
JA - IEICE TRANSACTIONS on Information
Y1 - April 2020
AB - The triaxial accelerometer is one of the most important sensors for human activity recognition (HAR). It has been observed that the relations between the axes of a triaxial accelerometer plays a significant role in improving the accuracy of activity recognition. However, the existing research rarely focuses on these relations, but rather on the fusion of multiple sensors. In this paper, we propose a data fusion-based convolutional neural network (CNN) approach to effectively use the relations between the axes. We design a single-channel data fusion method and multichannel data fusion method in consideration of the diversified formats of sensor data. After obtaining the fused data, a CNN is used to extract the features and perform classification. The experiments show that the proposed approach has an advantage over the CNN in accuracy. Moreover, the single-channel model achieves an accuracy of 98.83% with the WISDM dataset, which is higher than that of state-of-the-art methods.
ER -