Comparison of Activity Recognition using 2D and 3D Skeletal Joint Data

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

With the recent development of cheap and accurate depth sensors, activity recognition research has largely focused on the use of features created from 3D, rather than 2D, skeletal joint location. Nevertheless, conventional 2D RGB cameras remain an attractive data collection tool due to their low cost and ease of use. This study investigates the benefits of using 2D skeletal joints for activity recognition using visualisation and an exemplar classifier. Results show that 2D models can be as informative as 3D models, demonstrating the informativeness of joints extracted from RGB video.
LanguageEnglish
Title of host publicationIrish Machine Vision & Image Processing Conference proceedings
Subtitle of host publicationIMVIP 2019
Pages13
Number of pages20
ISBN (Electronic) ISBN 978-0-9934207-4-0
Publication statusPublished - Aug 2019
EventIrish Machine Vision & Image Processing IMVIP 2019 - Grangegorman Campus, Technological University, Dublin, Ireland
Duration: 28 Aug 201930 Aug 2019
http://www.imvip.ie/

Conference

ConferenceIrish Machine Vision & Image Processing IMVIP 2019
Abbreviated titleIMVIP
CountryIreland
CityDublin
Period28/08/1930/08/19
Internet address

Fingerprint

Classifiers
Visualization
Cameras
Sensors
Costs

Keywords

  • Activity Recognition
  • 3D Skeletal Joints
  • 2D Skeletal Joints
  • RGB

Cite this

Marshall, F., Zhang, S., & Scotney, B. (2019). Comparison of Activity Recognition using 2D and 3D Skeletal Joint Data. In Irish Machine Vision & Image Processing Conference proceedings: IMVIP 2019 (pp. 13)
Marshall, Fiona ; Zhang, Shuai ; Scotney, Bryan. / Comparison of Activity Recognition using 2D and 3D Skeletal Joint Data. Irish Machine Vision & Image Processing Conference proceedings: IMVIP 2019. 2019. pp. 13
@inproceedings{60b4541e643749018659e66435871687,
title = "Comparison of Activity Recognition using 2D and 3D Skeletal Joint Data",
abstract = "With the recent development of cheap and accurate depth sensors, activity recognition research has largely focused on the use of features created from 3D, rather than 2D, skeletal joint location. Nevertheless, conventional 2D RGB cameras remain an attractive data collection tool due to their low cost and ease of use. This study investigates the benefits of using 2D skeletal joints for activity recognition using visualisation and an exemplar classifier. Results show that 2D models can be as informative as 3D models, demonstrating the informativeness of joints extracted from RGB video.",
keywords = "Activity Recognition, 3D Skeletal Joints, 2D Skeletal Joints, RGB",
author = "Fiona Marshall and Shuai Zhang and Bryan Scotney",
year = "2019",
month = "8",
language = "English",
pages = "13",
booktitle = "Irish Machine Vision & Image Processing Conference proceedings",

}

Marshall, F, Zhang, S & Scotney, B 2019, Comparison of Activity Recognition using 2D and 3D Skeletal Joint Data. in Irish Machine Vision & Image Processing Conference proceedings: IMVIP 2019. pp. 13, Irish Machine Vision & Image Processing IMVIP 2019, Dublin, Ireland, 28/08/19.

Comparison of Activity Recognition using 2D and 3D Skeletal Joint Data. / Marshall, Fiona; Zhang, Shuai; Scotney, Bryan.

Irish Machine Vision & Image Processing Conference proceedings: IMVIP 2019. 2019. p. 13.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Comparison of Activity Recognition using 2D and 3D Skeletal Joint Data

AU - Marshall, Fiona

AU - Zhang, Shuai

AU - Scotney, Bryan

PY - 2019/8

Y1 - 2019/8

N2 - With the recent development of cheap and accurate depth sensors, activity recognition research has largely focused on the use of features created from 3D, rather than 2D, skeletal joint location. Nevertheless, conventional 2D RGB cameras remain an attractive data collection tool due to their low cost and ease of use. This study investigates the benefits of using 2D skeletal joints for activity recognition using visualisation and an exemplar classifier. Results show that 2D models can be as informative as 3D models, demonstrating the informativeness of joints extracted from RGB video.

AB - With the recent development of cheap and accurate depth sensors, activity recognition research has largely focused on the use of features created from 3D, rather than 2D, skeletal joint location. Nevertheless, conventional 2D RGB cameras remain an attractive data collection tool due to their low cost and ease of use. This study investigates the benefits of using 2D skeletal joints for activity recognition using visualisation and an exemplar classifier. Results show that 2D models can be as informative as 3D models, demonstrating the informativeness of joints extracted from RGB video.

KW - Activity Recognition

KW - 3D Skeletal Joints

KW - 2D Skeletal Joints

KW - RGB

UR - http://www.imvip.ie/

M3 - Conference contribution

SP - 13

BT - Irish Machine Vision & Image Processing Conference proceedings

ER -

Marshall F, Zhang S, Scotney B. Comparison of Activity Recognition using 2D and 3D Skeletal Joint Data. In Irish Machine Vision & Image Processing Conference proceedings: IMVIP 2019. 2019. p. 13