Slow feature analysis for human action recognition

Zhang, Z; Tao, D

Slow feature analysis for human action recognition

Zhang, Z Tao, D

Permalink

Publication Type:: Journal Article
Citation:: IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34 (3), pp. 436 - 450
Issue Date:: 2012-01-30

Closed Access

	Filename	Description	Size
	2011005510OK.pdf		5.49 MB	Adobe PDF	View/Open

Copyright Clearance Process

Recently Added
In Progress
Closed Access

This item is closed access and not available.

Full metadata record

Field	Value	Language
dc.contributor.author	Zhang, Z	en_US
dc.contributor.author	Tao, D https://orcid.org/0000-0001-7225-5449	en_US
dc.date.issued	2012-01-30	en_US
dc.identifier.citation	IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34 (3), pp. 436 - 450	en_US
dc.identifier.issn	0162-8828	en_US
dc.identifier.uri	http://hdl.handle.net/10453/22229
dc.description.abstract	Slow Feature Analysis (SFA) extracts slowly varying features from a quickly varying input signal [1]. It has been successfully applied to modeling the visual receptive fields of the cortical neurons. Sufficient experimental results in neuroscience suggest that the temporal slowness principle is a general learning principle in visual perception. In this paper, we introduce the SFA framework to the problem of human action recognition by incorporating the discriminative information with SFA learning and considering the spatial relationship of body parts. In particular, we consider four kinds of SFA learning strategies, including the original unsupervised SFA (U-SFA), the supervised SFA (S-SFA), the discriminative SFA (D-SFA), and the spatial discriminative SFA (SD-SFA), to extract slow feature functions from a large amount of training cuboids which are obtained by random sampling in motion boundaries. Afterward, to represent action sequences, the squared first order temporal derivatives are accumulated over all transformed cuboids into one feature vector, which is termed the Accumulated Squared Derivative (ASD) feature. The ASD feature encodes the statistical distribution of slow features in an action sequence. Finally, a linear support vector machine (SVM) is trained to classify actions represented by ASD features. We conduct extensive experiments, including two sets of control experiments, two sets of large scale experiments on the KTH and Weizmann databases, and two sets of experiments on the CASIA and UT-interaction databases, to demonstrate the effectiveness of SFA for human action recognition. Experimental results suggest that the SFA-based approach 1) is able to extract useful motion patterns and improves the recognition performance, 2) requires less intermediate processing steps but achieves comparable or even better performance, and 3) has good potential to recognize complex multiperson activities. © 2012 IEEE.	en_US
dc.relation.ispartof	IEEE Transactions on Pattern Analysis and Machine Intelligence	en_US
dc.relation.isbasedon	10.1109/TPAMI.2011.157	en_US
dc.subject.classification	Artificial Intelligence & Image Processing	en_US
dc.subject.mesh	Humans	en_US
dc.subject.mesh	Visual Perception	en_US
dc.subject.mesh	Algorithms	en_US
dc.subject.mesh	Support Vector Machines	en_US
dc.subject.mesh	Support Vector Machine	en_US
dc.title	Slow feature analysis for human action recognition	en_US
dc.type	Journal Article
utslib.citation.volume	3	en_US
utslib.citation.volume	34	en_US
utslib.for	0801 Artificial Intelligence and Image Processing	en_US
utslib.for	0806 Information Systems	en_US
utslib.for	0906 Electrical and Electronic Engineering	en_US
pubs.embargo.period	Not known	en_US
pubs.organisational-group	/University of Technology Sydney
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology
utslib.copyright.status	closed_access
pubs.issue	3	en_US
pubs.publication-status	Published	en_US
pubs.volume	34	en_US

Abstract:

Slow Feature Analysis (SFA) extracts slowly varying features from a quickly varying input signal [1]. It has been successfully applied to modeling the visual receptive fields of the cortical neurons. Sufficient experimental results in neuroscience suggest that the temporal slowness principle is a general learning principle in visual perception. In this paper, we introduce the SFA framework to the problem of human action recognition by incorporating the discriminative information with SFA learning and considering the spatial relationship of body parts. In particular, we consider four kinds of SFA learning strategies, including the original unsupervised SFA (U-SFA), the supervised SFA (S-SFA), the discriminative SFA (D-SFA), and the spatial discriminative SFA (SD-SFA), to extract slow feature functions from a large amount of training cuboids which are obtained by random sampling in motion boundaries. Afterward, to represent action sequences, the squared first order temporal derivatives are accumulated over all transformed cuboids into one feature vector, which is termed the Accumulated Squared Derivative (ASD) feature. The ASD feature encodes the statistical distribution of slow features in an action sequence. Finally, a linear support vector machine (SVM) is trained to classify actions represented by ASD features. We conduct extensive experiments, including two sets of control experiments, two sets of large scale experiments on the KTH and Weizmann databases, and two sets of experiments on the CASIA and UT-interaction databases, to demonstrate the effectiveness of SFA for human action recognition. Experimental results suggest that the SFA-based approach 1) is able to extract useful motion patterns and improves the recognition performance, 2) requires less intermediate processing steps but achieves comparable or even better performance, and 3) has good potential to recognize complex multiperson activities. © 2012 IEEE.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/22229