Please use this identifier to cite or link to this item: http://hdl.handle.net/10316/27816
Title: Trajectory-based human action segmentation
Authors: Santos, Luís 
Khoshhal, Kamrad 
Dias, Jorge 
Keywords: Motion segmentation; Classification framework; Signal processing; Motion variability; Adaptive sliding window
Issue Date: Feb-2015
Publisher: Elsevier
Keywords: Motion segmentation; Classification framework; Signal processing; Motion variability; Adaptive sliding window
Issue Date: Feb-2015
Publisher: Elsevier
Citation: SANTOS, Luís; KHOSHHAL, Kamrad; DIAS, Jorge - Trajectory-based human action segmentation. "Pattern Recognition". ISSN 0031-3203. Vol. 48 Nº. 2 (2015) p. 568–579
Abstract: This paper proposes a sliding window approach, whose length and time shift are dynamically adaptable in order to improve model confidence, speed and segmentation accuracy in human action sequences. Activity recognition is the process of inferring an action class from a set of observations acquired by sensors. We address the temporal segmentation problem of body part trajectories in Cartesian Space in which features are generated using Discrete Fast Fourier Transform (DFFT) and Power Spectrum (PS). We pose this as an entropy minimization problem. Using entropy from the classifier output as a feedback parameter, we continuously adjust the two key parameters in a sliding window approach, to maximize the model confidence at every step. The proposed classifier is a Dynamic Bayesian Network (DBN) model where classes are estimated using Bayesian inference. We compare our approach with our previously developed fixed window method. Experiments show that our method accurately recognizes and segments activities, with improved model confidence and faster convergence times, exhibiting anticipatory capabilities. Our work demonstrates that entropy feedback mitigates variability problems, and our method is applicable in research areas where action segmentation and classification is used. A working demo source code is provided online for academical dissemination purposes, by requesting the authors.
URI: http://hdl.handle.net/10316/27816
ISSN: 0031-3203
DOI: 10.1016/j.patcog.2014.08.015
Rights: openAccess
Appears in Collections:I&D ISR - Artigos em Revistas Internacionais
FCTUC Eng.Electrotécnica - Artigos em Revistas Internacionais

Files in This Item:
File Description SizeFormat
Trajectory-based human action segmentation.pdf6.89 MBAdobe PDFView/Open
Show full item record

SCOPUSTM   
Citations

16
checked on Jun 25, 2019

WEB OF SCIENCETM
Citations

11
checked on Jun 25, 2019

Page view(s) 5

1,326
checked on Aug 13, 2019

Download(s) 50

410
checked on Aug 13, 2019

Google ScholarTM

Check

Altmetric

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.