Time-frequency based features for classification of walking patterns

Ronny K. Ibrahim; Eliathamby Ambikairajah; Branko G. Celler; Nigel H. Lovell

doi:10.1109/ICDSP.2007.4288550

Time-frequency based features for classification of walking patterns

Ronny K. Ibrahim, Eliathamby Ambikairajah, Branko G. Celler, Nigel H. Lovell

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

27 Citations (Scopus)

Abstract

The analysis of gait data has been a challenging problem and several new approaches have been proposed in recent years. This paper describes a novel front-end for classification of gait patterns using data obtained from a tri-axial accelerometer. The novel features consist of delta features, low and high frequency signal variations and energy variations in both frequency bands. The back-end of the system is a Gaussian mixture model based classifier. Using Bayesian adaptation, an overall classification accuracy of 96.1% was achieved for five walking patterns.

Original language	English
Title of host publication	2007 15th International Conference on Digital Signal Processing, DSP 2007
Subtitle of host publication	1-4 July 2007, Cardiff University, Wales, UK
Editors	Saeid Sanei, Jonathon A. Chambers, John McWhirter, Yulia Hicks, Anthony G. Constantinides
Place of Publication	Piscataway, NJ
Publisher	Institute of Electrical and Electronics Engineers (IEEE)
Pages	187-190
Number of pages	4
ISBN (Print)	1424408822, 9781424408825
DOIs	https://doi.org/10.1109/ICDSP.2007.4288550
Publication status	Published - 2007
Event	15th International Conference on Digital Signal Processing - Cardiff Duration: 1 Jul 2007 → 4 Jul 2007

Conference

Conference	15th International Conference on Digital Signal Processing
City	Cardiff
Period	1/07/07 → 4/07/07

Keywords

gait patterns
accelerometry
ambulatory monitoring
Gaussian mixture models

Access to Document

10.1109/ICDSP.2007.4288550

Cite this

Ibrahim, R. K., Ambikairajah, E., Celler, B. G., & Lovell, N. H. (2007). Time-frequency based features for classification of walking patterns. In S. Sanei, J. A. Chambers, J. McWhirter, Y. Hicks, & A. G. Constantinides (Eds.), 2007 15th International Conference on Digital Signal Processing, DSP 2007: 1-4 July 2007, Cardiff University, Wales, UK (pp. 187-190). Article 4288550 Institute of Electrical and Electronics Engineers (IEEE). https://doi.org/10.1109/ICDSP.2007.4288550

Ibrahim, Ronny K. ; Ambikairajah, Eliathamby ; Celler, Branko G. et al. / Time-frequency based features for classification of walking patterns. 2007 15th International Conference on Digital Signal Processing, DSP 2007: 1-4 July 2007, Cardiff University, Wales, UK. editor / Saeid Sanei ; Jonathon A. Chambers ; John McWhirter ; Yulia Hicks ; Anthony G. Constantinides. Piscataway, NJ : Institute of Electrical and Electronics Engineers (IEEE), 2007. pp. 187-190

@inproceedings{5cf878f0d5594792b8043460c6737e61,

title = "Time-frequency based features for classification of walking patterns",

abstract = "The analysis of gait data has been a challenging problem and several new approaches have been proposed in recent years. This paper describes a novel front-end for classification of gait patterns using data obtained from a tri-axial accelerometer. The novel features consist of delta features, low and high frequency signal variations and energy variations in both frequency bands. The back-end of the system is a Gaussian mixture model based classifier. Using Bayesian adaptation, an overall classification accuracy of 96.1\% was achieved for five walking patterns.",

keywords = "gait patterns, accelerometry, ambulatory monitoring, Gaussian mixture models",

author = "Ibrahim, \{Ronny K.\} and Eliathamby Ambikairajah and Celler, \{Branko G.\} and Lovell, \{Nigel H.\}",

year = "2007",

doi = "10.1109/ICDSP.2007.4288550",

language = "English",

isbn = "1424408822",

pages = "187--190",

editor = "Saeid Sanei and Chambers, \{Jonathon A.\} and John McWhirter and Yulia Hicks and Constantinides, \{Anthony G.\}",

booktitle = "2007 15th International Conference on Digital Signal Processing, DSP 2007",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

address = "United States",

note = "15th International Conference on Digital Signal Processing ; Conference date: 01-07-2007 Through 04-07-2007",

}

Ibrahim, RK, Ambikairajah, E, Celler, BG & Lovell, NH 2007, Time-frequency based features for classification of walking patterns. in S Sanei, JA Chambers, J McWhirter, Y Hicks & AG Constantinides (eds), 2007 15th International Conference on Digital Signal Processing, DSP 2007: 1-4 July 2007, Cardiff University, Wales, UK., 4288550, Institute of Electrical and Electronics Engineers (IEEE), Piscataway, NJ, pp. 187-190, 15th International Conference on Digital Signal Processing, Cardiff, 1/07/07. https://doi.org/10.1109/ICDSP.2007.4288550

Time-frequency based features for classification of walking patterns. / Ibrahim, Ronny K.; Ambikairajah, Eliathamby; Celler, Branko G. et al.
2007 15th International Conference on Digital Signal Processing, DSP 2007: 1-4 July 2007, Cardiff University, Wales, UK. ed. / Saeid Sanei; Jonathon A. Chambers; John McWhirter; Yulia Hicks; Anthony G. Constantinides. Piscataway, NJ: Institute of Electrical and Electronics Engineers (IEEE), 2007. p. 187-190 4288550.

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

TY - GEN

T1 - Time-frequency based features for classification of walking patterns

AU - Ibrahim, Ronny K.

AU - Ambikairajah, Eliathamby

AU - Celler, Branko G.

AU - Lovell, Nigel H.

PY - 2007

Y1 - 2007

N2 - The analysis of gait data has been a challenging problem and several new approaches have been proposed in recent years. This paper describes a novel front-end for classification of gait patterns using data obtained from a tri-axial accelerometer. The novel features consist of delta features, low and high frequency signal variations and energy variations in both frequency bands. The back-end of the system is a Gaussian mixture model based classifier. Using Bayesian adaptation, an overall classification accuracy of 96.1% was achieved for five walking patterns.

AB - The analysis of gait data has been a challenging problem and several new approaches have been proposed in recent years. This paper describes a novel front-end for classification of gait patterns using data obtained from a tri-axial accelerometer. The novel features consist of delta features, low and high frequency signal variations and energy variations in both frequency bands. The back-end of the system is a Gaussian mixture model based classifier. Using Bayesian adaptation, an overall classification accuracy of 96.1% was achieved for five walking patterns.

KW - gait patterns

KW - accelerometry

KW - ambulatory monitoring

KW - Gaussian mixture models

UR - https://www.scopus.com/pages/publications/47649084592

U2 - 10.1109/ICDSP.2007.4288550

DO - 10.1109/ICDSP.2007.4288550

M3 - Conference proceeding contribution

AN - SCOPUS:47649084592

SN - 1424408822

SN - 9781424408825

SP - 187

EP - 190

BT - 2007 15th International Conference on Digital Signal Processing, DSP 2007

A2 - Sanei, Saeid

A2 - Chambers, Jonathon A.

A2 - McWhirter, John

A2 - Hicks, Yulia

A2 - Constantinides, Anthony G.

PB - Institute of Electrical and Electronics Engineers (IEEE)

CY - Piscataway, NJ

T2 - 15th International Conference on Digital Signal Processing

Y2 - 1 July 2007 through 4 July 2007

ER -

Ibrahim RK, Ambikairajah E, Celler BG, Lovell NH. Time-frequency based features for classification of walking patterns. In Sanei S, Chambers JA, McWhirter J, Hicks Y, Constantinides AG, editors, 2007 15th International Conference on Digital Signal Processing, DSP 2007: 1-4 July 2007, Cardiff University, Wales, UK. Piscataway, NJ: Institute of Electrical and Electronics Engineers (IEEE). 2007. p. 187-190. 4288550 doi: 10.1109/ICDSP.2007.4288550

Time-frequency based features for classification of walking patterns

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this