Multi-modality fusion of floor and ambulatory sensors for gait classification

Syed Usama YUNAS; Abdullah Alharthi; Krikor Ozanyan

doi:10.1109/ISIE.2019.8781127

Multi-modality fusion of floor and ambulatory sensors for gait classification

Syed Usama YUNAS, Abdullah Alharthi, Krikor Ozanyan

Research output: Contribution to conference › Paper › peer-review

239 Downloads (Pure)

Abstract

In a case study of gait classification from floor and ambulatory sensors, we compare results with data from each modality. The automatic extraction of features is achieved by Principle Component Analysis and Canonical Correlation Analysis, the latter performing better even with a reduced number of components used. Non-linear classifiers are most efficient for fused features. With a Kernel Support Vector Machine around 94% accuracy is demonstrated, improving over the 87% and 79% accuracies obtained with separate floor and ambulatory sensor data, respectively.

Original language	English
Pages	1467-1472
Number of pages	6
DOIs	https://doi.org/10.1109/ISIE.2019.8781127
Publication status	Published - 1 Aug 2019
Event	2019 IEEE 28th International Symposium on Industrial Electronics (ISIE) - Canada, Vancouver, Canada Duration: 12 Jun 2019 → 14 Jun 2019

Conference

Conference	2019 IEEE 28th International Symposium on Industrial Electronics (ISIE)
Country/Territory	Canada
City	Vancouver
Period	12/06/19 → 14/06/19

Keywords

floor sensors, wearable sensors, sensor fusion, machine learning (ML), inertial measurement unit (IMU), principal component analysis (PCA), canonical correlation analysis (CCA)

Access to Document

10.1109/ISIE.2019.8781127

Multi-modality fusion of floor and ambulatory sensors for gait classification

Cite this

@conference{4b4394b3def8499188a5ed431143c89a,

title = "Multi-modality fusion of floor and ambulatory sensors for gait classification",

abstract = "In a case study of gait classification from floor and ambulatory sensors, we compare results with data from each modality. The automatic extraction of features is achieved by Principle Component Analysis and Canonical Correlation Analysis, the latter performing better even with a reduced number of components used. Non-linear classifiers are most efficient for fused features. With a Kernel Support Vector Machine around 94\% accuracy is demonstrated, improving over the 87\% and 79\% accuracies obtained with separate floor and ambulatory sensor data, respectively.",

keywords = "floor sensors, wearable sensors, sensor fusion, machine learning (ML), inertial measurement unit (IMU), principal component analysis (PCA), canonical correlation analysis (CCA)",

author = "YUNAS, \{Syed Usama\} and Abdullah Alharthi and Krikor Ozanyan",

year = "2019",

month = aug,

day = "1",

doi = "10.1109/ISIE.2019.8781127",

language = "English",

pages = "1467--1472",

note = " 2019 IEEE 28th International Symposium on Industrial Electronics (ISIE) ; Conference date: 12-06-2019 Through 14-06-2019",

}

TY - CONF

T1 - Multi-modality fusion of floor and ambulatory sensors for gait classification

AU - YUNAS, Syed Usama

AU - Alharthi, Abdullah

AU - Ozanyan, Krikor

PY - 2019/8/1

Y1 - 2019/8/1

N2 - In a case study of gait classification from floor and ambulatory sensors, we compare results with data from each modality. The automatic extraction of features is achieved by Principle Component Analysis and Canonical Correlation Analysis, the latter performing better even with a reduced number of components used. Non-linear classifiers are most efficient for fused features. With a Kernel Support Vector Machine around 94% accuracy is demonstrated, improving over the 87% and 79% accuracies obtained with separate floor and ambulatory sensor data, respectively.

AB - In a case study of gait classification from floor and ambulatory sensors, we compare results with data from each modality. The automatic extraction of features is achieved by Principle Component Analysis and Canonical Correlation Analysis, the latter performing better even with a reduced number of components used. Non-linear classifiers are most efficient for fused features. With a Kernel Support Vector Machine around 94% accuracy is demonstrated, improving over the 87% and 79% accuracies obtained with separate floor and ambulatory sensor data, respectively.

KW - floor sensors, wearable sensors, sensor fusion, machine learning (ML), inertial measurement unit (IMU), principal component analysis (PCA), canonical correlation analysis (CCA)

UR - https://www.scopus.com/pages/publications/85070614156

U2 - 10.1109/ISIE.2019.8781127

DO - 10.1109/ISIE.2019.8781127

M3 - Paper

SP - 1467

EP - 1472

T2 - 2019 IEEE 28th International Symposium on Industrial Electronics (ISIE)

Y2 - 12 June 2019 through 14 June 2019

ER -

Multi-modality fusion of floor and ambulatory sensors for gait classification

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this