Spatio-temporal event classification using time-series kernel based structured sparsity

László A. Jeni, András Lorincz, Zoltán Szabó, Jeffrey F. Cohn, Takeo Kanade

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

16 Citations (Scopus)

Abstract

In many behavioral domains, such as facial expression and gesture, sparse structure is prevalent. This sparsity would be well suited for event detection but for one problem. Features typically are confounded by alignment error in space and time. As a consequence, high-dimensional representations such as SIFT and Gabor features have been favored despite their much greater computational cost and potential loss of information. We propose a Kernel Structured Sparsity (KSS) method that can handle both the temporal alignment problem and the structured sparse reconstruction within a common framework, and it can rely on simple features. We characterize spatio-temporal events as time-series of motion patterns and by utilizing time-series kernels we apply standard structured-sparse coding techniques to tackle this important problem. We evaluated the KSS method using both gesture and facial expression datasets that include spontaneous behavior and differ in degree of difficulty and type of ground truth coding. KSS outperformed both sparse and non-sparse methods that utilize complex image features and their temporal extensions. In the case of early facial event classification KSS had 10% higher accuracy as measured by F 1 score over kernel SVM methods.

Original languageEnglish
Title of host publicationComputer Vision, ECCV 2014 - 13th European Conference, Proceedings
PublisherSpringer Verlag
Pages135-150
Number of pages16
EditionPART 4
ISBN (Print)9783319105925
DOIs
Publication statusPublished - Jan 1 2014
Event13th European Conference on Computer Vision, ECCV 2014 - Zurich, Switzerland
Duration: Sep 6 2014Sep 12 2014

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
NumberPART 4
Volume8692 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other13th European Conference on Computer Vision, ECCV 2014
CountrySwitzerland
CityZurich
Period9/6/149/12/14

Keywords

  • facial expression classification
  • gesture recognition
  • structured sparsity
  • time-series kernels

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Spatio-temporal event classification using time-series kernel based structured sparsity'. Together they form a unique fingerprint.

  • Cite this

    Jeni, L. A., Lorincz, A., Szabó, Z., Cohn, J. F., & Kanade, T. (2014). Spatio-temporal event classification using time-series kernel based structured sparsity. In Computer Vision, ECCV 2014 - 13th European Conference, Proceedings (PART 4 ed., pp. 135-150). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8692 LNCS, No. PART 4). Springer Verlag. https://doi.org/10.1007/978-3-319-10593-2_10