From Kinect skeleton data to hand gesture recognition with radar
In an era where man-machine interaction increasingly uses remote sensing, gesture recognition through use of the micro-Doppler (mD) effect is an emerging application which has attracted great interest. It is a sensible solution and here the authors show its potential for detecting aperiodic human mo...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Wiley
2019-09-01
|
Series: | The Journal of Engineering |
Subjects: | |
Online Access: | https://digital-library.theiet.org/content/journals/10.1049/joe.2019.0557 |
id |
doaj-b0ec34894eb14d96a50410893fd7267c |
---|---|
record_format |
Article |
spelling |
doaj-b0ec34894eb14d96a50410893fd7267c2021-04-02T11:48:59ZengWileyThe Journal of Engineering2051-33052019-09-0110.1049/joe.2019.0557JOE.2019.0557From Kinect skeleton data to hand gesture recognition with radarJiayi Li0Aman Shrestha1Julien Le Kernec2Francesco Fioranelli3Francesco Fioranelli4School of Information and Electronics, University of Electronic Science and Technology of ChinaCommunicaiton, Sensing and Imaging Group, School of Engineering, University of GlasgowSchool of Information and Electronics, University of Electronic Science and Technology of ChinaCommunicaiton, Sensing and Imaging Group, School of Engineering, University of GlasgowCommunicaiton, Sensing and Imaging Group, School of Engineering, University of GlasgowIn an era where man-machine interaction increasingly uses remote sensing, gesture recognition through use of the micro-Doppler (mD) effect is an emerging application which has attracted great interest. It is a sensible solution and here the authors show its potential for detecting aperiodic human movements. In this study, the authors classify ten hand gestures with a set of handcrafted features using simulated mD signatures generated from Kinect skeleton data. Data augmentation in the form of synthetic minority oversampling technique has been applied to create synthetic samples and classified with the support vector machine and K-nearest neighbour classifier with classification rate of 71.1 and 51% achieved. Finally, using weights generated by an action pair based one vs. one classification layer improves classification accuracy by 24.7 and 28.4%.https://digital-library.theiet.org/content/journals/10.1049/joe.2019.0557support vector machinesimage classificationgesture recognitionimage sensorsfeature extractionimage samplingradar imagingdoppler radaraperiodic human movementshand gestureshandcrafted featuressimulated md signatureskinect skeleton datadata augmentationsynthetic minority oversampling techniquesynthetic samplessupport vector machineneighbour classifiergesture recognitionman-machine interactionremote sensingmicrodoppler effect |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Jiayi Li Aman Shrestha Julien Le Kernec Francesco Fioranelli Francesco Fioranelli |
spellingShingle |
Jiayi Li Aman Shrestha Julien Le Kernec Francesco Fioranelli Francesco Fioranelli From Kinect skeleton data to hand gesture recognition with radar The Journal of Engineering support vector machines image classification gesture recognition image sensors feature extraction image sampling radar imaging doppler radar aperiodic human movements hand gestures handcrafted features simulated md signatures kinect skeleton data data augmentation synthetic minority oversampling technique synthetic samples support vector machine neighbour classifier gesture recognition man-machine interaction remote sensing microdoppler effect |
author_facet |
Jiayi Li Aman Shrestha Julien Le Kernec Francesco Fioranelli Francesco Fioranelli |
author_sort |
Jiayi Li |
title |
From Kinect skeleton data to hand gesture recognition with radar |
title_short |
From Kinect skeleton data to hand gesture recognition with radar |
title_full |
From Kinect skeleton data to hand gesture recognition with radar |
title_fullStr |
From Kinect skeleton data to hand gesture recognition with radar |
title_full_unstemmed |
From Kinect skeleton data to hand gesture recognition with radar |
title_sort |
from kinect skeleton data to hand gesture recognition with radar |
publisher |
Wiley |
series |
The Journal of Engineering |
issn |
2051-3305 |
publishDate |
2019-09-01 |
description |
In an era where man-machine interaction increasingly uses remote sensing, gesture recognition through use of the micro-Doppler (mD) effect is an emerging application which has attracted great interest. It is a sensible solution and here the authors show its potential for detecting aperiodic human movements. In this study, the authors classify ten hand gestures with a set of handcrafted features using simulated mD signatures generated from Kinect skeleton data. Data augmentation in the form of synthetic minority oversampling technique has been applied to create synthetic samples and classified with the support vector machine and K-nearest neighbour classifier with classification rate of 71.1 and 51% achieved. Finally, using weights generated by an action pair based one vs. one classification layer improves classification accuracy by 24.7 and 28.4%. |
topic |
support vector machines image classification gesture recognition image sensors feature extraction image sampling radar imaging doppler radar aperiodic human movements hand gestures handcrafted features simulated md signatures kinect skeleton data data augmentation synthetic minority oversampling technique synthetic samples support vector machine neighbour classifier gesture recognition man-machine interaction remote sensing microdoppler effect |
url |
https://digital-library.theiet.org/content/journals/10.1049/joe.2019.0557 |
work_keys_str_mv |
AT jiayili fromkinectskeletondatatohandgesturerecognitionwithradar AT amanshrestha fromkinectskeletondatatohandgesturerecognitionwithradar AT julienlekernec fromkinectskeletondatatohandgesturerecognitionwithradar AT francescofioranelli fromkinectskeletondatatohandgesturerecognitionwithradar AT francescofioranelli fromkinectskeletondatatohandgesturerecognitionwithradar |
_version_ |
1721571206837043200 |