Using fuzzy gaussian inference and genetic programming to classify 3D human motions

Mehdi Khoury; Honghai Liu

doi:10.1007/978-1-84996-329-9_5

Using fuzzy gaussian inference and genetic programming to classify 3D human motions

Mehdi Khoury^*
, Honghai Liu

^*Corresponding author for this work

University of Portsmouth

Research output: Chapter in Book/Report/Conference proceeding › Chapter › peer-review

Abstract

This research introduces and builds on the concept of Fuzzy Gaussian Inference (FGI) (Khoury and Liu in Proceedings of UKCI, 2008 and IEEE Workshop on Robotic Intelligence in Informationally Structured Space (RiiSS 2009), 2009) as a novel way to build Fuzzy Membership Functions that map to hidden Probability Distributions underlying human motions. This method is now combined with a Genetic Programming Fuzzy rule-based system in order to classify boxing moves from natural human Motion Capture data. In this experiment, FGI alone is able to recognise seven different boxing stances simultaneously with an accuracy superior to a GMM-based classifier. Results seem to indicate that adding an evolutionary Fuzzy Inference Engine on top of FGI improves the accuracy of the classifier in a consistent way.

Original language	English
Title of host publication	Advanced Information and Knowledge Processing
Publisher	Springer-Verlag London Ltd
Pages	95-116
Number of pages	22
DOIs	https://doi.org/10.1007/978-1-84996-329-9_5
State	Published - 2010
Externally published	Yes

Publication series

Name	Advanced Information and Knowledge Processing
Volume	54
ISSN (Print)	1610-3947
ISSN (Electronic)	2197-8441

Access to Document

10.1007/978-1-84996-329-9_5

Cite this

@inbook{2b1bc6cc16b748aeb16f7ad3eff8e118,

title = "Using fuzzy gaussian inference and genetic programming to classify 3D human motions",

abstract = "This research introduces and builds on the concept of Fuzzy Gaussian Inference (FGI) (Khoury and Liu in Proceedings of UKCI, 2008 and IEEE Workshop on Robotic Intelligence in Informationally Structured Space (RiiSS 2009), 2009) as a novel way to build Fuzzy Membership Functions that map to hidden Probability Distributions underlying human motions. This method is now combined with a Genetic Programming Fuzzy rule-based system in order to classify boxing moves from natural human Motion Capture data. In this experiment, FGI alone is able to recognise seven different boxing stances simultaneously with an accuracy superior to a GMM-based classifier. Results seem to indicate that adding an evolutionary Fuzzy Inference Engine on top of FGI improves the accuracy of the classifier in a consistent way.",

author = "Mehdi Khoury and Honghai Liu",

note = "Publisher Copyright: {\textcopyright} Springer-Verlag London Limited 2010.",

year = "2010",

doi = "10.1007/978-1-84996-329-9\_5",

language = "英语",

series = "Advanced Information and Knowledge Processing",

publisher = "Springer-Verlag London Ltd",

pages = "95--116",

booktitle = "Advanced Information and Knowledge Processing",

}

TY - CHAP

T1 - Using fuzzy gaussian inference and genetic programming to classify 3D human motions

AU - Khoury, Mehdi

AU - Liu, Honghai

PY - 2010

Y1 - 2010

N2 - This research introduces and builds on the concept of Fuzzy Gaussian Inference (FGI) (Khoury and Liu in Proceedings of UKCI, 2008 and IEEE Workshop on Robotic Intelligence in Informationally Structured Space (RiiSS 2009), 2009) as a novel way to build Fuzzy Membership Functions that map to hidden Probability Distributions underlying human motions. This method is now combined with a Genetic Programming Fuzzy rule-based system in order to classify boxing moves from natural human Motion Capture data. In this experiment, FGI alone is able to recognise seven different boxing stances simultaneously with an accuracy superior to a GMM-based classifier. Results seem to indicate that adding an evolutionary Fuzzy Inference Engine on top of FGI improves the accuracy of the classifier in a consistent way.

AB - This research introduces and builds on the concept of Fuzzy Gaussian Inference (FGI) (Khoury and Liu in Proceedings of UKCI, 2008 and IEEE Workshop on Robotic Intelligence in Informationally Structured Space (RiiSS 2009), 2009) as a novel way to build Fuzzy Membership Functions that map to hidden Probability Distributions underlying human motions. This method is now combined with a Genetic Programming Fuzzy rule-based system in order to classify boxing moves from natural human Motion Capture data. In this experiment, FGI alone is able to recognise seven different boxing stances simultaneously with an accuracy superior to a GMM-based classifier. Results seem to indicate that adding an evolutionary Fuzzy Inference Engine on top of FGI improves the accuracy of the classifier in a consistent way.

UR - https://www.scopus.com/pages/publications/84950313695

U2 - 10.1007/978-1-84996-329-9_5

DO - 10.1007/978-1-84996-329-9_5

M3 - 章节

AN - SCOPUS:84950313695

T3 - Advanced Information and Knowledge Processing

SP - 95

EP - 116

BT - Advanced Information and Knowledge Processing

PB - Springer-Verlag London Ltd

ER -

Using fuzzy gaussian inference and genetic programming to classify 3D human motions

Abstract

Publication series

Access to Document

Other files and links

Fingerprint

Cite this