approved SFA « CIOP

SFA

SFA-TK Logo

project image SFA

Description

Slow Feature Analysis (SFA) is a technique invented by Laurenz Wiskott which finds unsupervisedly features in complex timeseries. The guidance principle of SFA is to find such combinations of the original inputs which vary slowly in time. The application areas of SFA range from visual pattern recognition over general classification up to unsupervised driving force detection in time series.

Within the project SFA @ CIOP (as part of the SOMA project), we have currently the following achievements:

  • We used SFA for the first time for gesture recognition [Koch10a] and achieved results comparable to other state-of-the-art gesture classifiers.
  • In order to do so, we expanded the original open-source, MATLAB-based SFA-Toolkit (sfa-tk V1.0) by Pietro Berkes to a new version sfa-tk V2.6 which is available for download. The extensions of V2.6 are:  more robust SVD-based SFA (see [Kon09b] for details), support for classification, support for cross validation, Gaussian classifier, some classification demos and sample data (UCI, gesture).
  • If the number of training samples is small, the original SFA classification algorithm will fail. We extended the SFA classification algorithm of Pietro Berkes by a parametric bootstrap algorithm which makes SFA classification applicable for small numbers of examples. This extension is available in sfa-tk V2.8 for download.
  • In a driving-force experiment we showed that SFA can detect signals slower than the driving force and that SFA may ‘switch concepts’ in a certain way which has some parallels to human perception [Kon09a][KonK10a][KonK10b].
  • In 2012 a group of students around Martin Zaefferer successfully ported sfa-tk V2.8 to R: The R-package rSFA is  available for download on CRAN. This allows to use SFA for unsupervised feature generation in data mining tasks and to integrate it as a general preprocessing tool in the SOMA project and the TDMR framework.

In the near future, we plan to pursue these further goals with SFA @ CIOP:

  • Improve gesture classification for the difficult task of recognizing gestures of unseen persons and towards even fewer training examples to learn new gestures.
  • Currently students are working on porting the MATLAB-based SFA-toolkit and the gesture recognition engine to the iPhone environment to make the gesture recognition available as iPhone App.

People

Wolfgang Konen, Kristine Hein, Patrick Koch

External links

Laurenz Wiskott et al. (2011), Slow feature analysis, Scholarpedia, 6(4):5282.

P.Berkes (2003), sfa-tk: Slow Feature Analysis Toolkit for Matlab (v.1.0.1). Or alternate site at ITB.

Publications

Publications in the area of Slow Feature Analysis (SFA) from the CIOP group:

  • [Kone11e (2011)] bibtex Go to document
    W. Konen, "Der SFA-Algorithmus für Klassifikation," Research Center CIOP (Computational Intelligence, Optimization and Data Mining), Cologne University of Applied Science, Faculty of Computer Science and Engineering Science, CIOP Technical Report 08/11, 2011.
    @TECHREPORT{Kone11e,
      author = {Wolfgang Konen},
      title = {Der SFA-Algorithmus f{\"u}r Klassifikation},
      institution = {Research Center CIOP (Computational Intelligence, Optimization and Data Mining)},
      year = {2011},
      type = {{CIOP} Technical Report},
      number = {08/11},
      address = {Cologne University of Applied Science, Faculty of Computer Science and Engineering Science},
      month = {November},
      file = {:http\://maanvs03.gm.fh-koeln.de/webpub/CIOPReports.d/Konen11e.d/Konen11e.pdf:PDF},
      issn = {2191-365X},
      keywords = {FiwaSoma SFA},
      owner = {Jrg},
      timestamp = {2011.11.29},
      url = {http://maanvs03.gm.fh-koeln.de/webpub/CIOPReports.d/Konen11e.d/Konen11e.pdf}
    }
  • [Kone11f (2011)] bibtex Go to document
    W. Konen, "SFA classification with few training data: Improvements with parametric bootstrap," Research Center CIOP (Computational Intelligence, Optimization and Data Mining), Cologne University of Applied Science, Faculty of Computer Science and Engineering Science, CIOP Technical Report 09/11, 2011.
    @TECHREPORT{Kone11f,
      author = {Wolfgang Konen},
      title = {SFA classification with few training data: Improvements with parametric bootstrap},
      institution = {Research Center CIOP (Computational Intelligence, Optimization and Data Mining)},
      year = {2011},
      type = {{CIOP} Technical Report},
      number = {09/11},
      address = {Cologne University of Applied Science, Faculty of Computer Science and Engineering Science},
      month = {November},
      file = {:http\://maanvs03.gm.fh-koeln.de/webpub/CIOPReports.d/Konen11f.d/Konen11f.pdf:PDF},
      issn = {2191-365X},
      keywords = {FiwaSoma SFA},
      owner = {Jrg},
      timestamp = {2011.11.29},
      url = {http://maanvs03.gm.fh-koeln.de/webpub/CIOPReports.d/Konen11f.d/Konen11f.pdf}
    }
  • [KonK10b (2011)] bibtex Go to document
    W. Konen and P. Koch, "The slowness principle: SFA can detect different slow components in nonstationary time series," International Journal of Innovative Computing and Applications (IJICA), vol. 3, iss. 1, pp. 3-10, 2011.
    @ARTICLE{KonK10b,
      author = {Wolfgang Konen and Patrick Koch},
      title = {The slowness principle: {SFA} can detect different slow components in nonstationary time series},
      journal = {International Journal of Innovative Computing and Applications (IJICA)},
      year = {2011},
      volume = {3},
      pages = {3--10},
      number = {1},
      doi = {10.1504/IJICA.2011.037946},
      keywords = {FiwaSoma SFA},
      owner = {wolfgang},
      timestamp = {2010.10.16},
      url = {http://www.gm.fh-koeln.de/~konen/Publikationen/IJICA2010-howslow.pdf}
    }
  • [KonK10a (2010)] bibtex Go to document
    W. Konen and P. Koch, "How slow is slow? SFA detects signals that are slower than the driving force," in Proc. 4th Int. Conf. on Bioinspired Optimization Methods and their Applications, BIOMA, Ljubljana, Slovenia, 2010.
    @INPROCEEDINGS{KonK10a,
      author = {Wolfgang Konen and Patrick Koch},
      title = {How slow is slow? {SFA} detects signals that are slower than the driving force},
      booktitle = {Proc. 4th Int. Conf. on Bioinspired Optimization Methods and their Applications, BIOMA, Ljubljana, Slovenia},
      year = {2010},
      editor = {B. Filipic, J. Silc},
      keywords = {FiwaSoma SFA},
      owner = {wolfgang},
      timestamp = {2010.10.16},
      url = {http://www.gm.fh-koeln.de/~konen/Publikationen/BIOMA10-howslow.pdf}
    }
  • [Hein10b (2010)] bibtex Go to document
    K. Hein, "Lernende Klassifikation beschleunigungsbasierter 3D-Gesten des Wii-Controllers," Research Center CIOP (Computational Intelligence, Optimization and Data Mining), Cologne University of Applied Science, Faculty of Computer Scienc and Engineering Science, CIOP Technical Report 01-10, 2010.
    @TECHREPORT{Hein10b,
      author = {Kristine Hein},
      title = {Lernende Klassifikation beschleunigungsbasierter 3D-Gesten des Wii-Controllers},
      institution = {Research Center CIOP (Computational Intelligence, Optimization and Data Mining)},
      year = {2010},
      type = {{CIOP} Technical Report},
      number = {01-10},
      address = {Cologne University of Applied Science, Faculty of Computer Scienc and Engineering Science},
      month = {Jan},
      file = {:http\://maanvs03.gm.fh-koeln.de/webpub/CIOPReports.d/Hein10b.d/Klassifikation3D.pdf:PDF},
      issn = {2191-365X},
      keywords = {FiwaSoma SFA},
      owner = {jstork},
      timestamp = {2011.01.21},
      url = {http://maanvs03.gm.fh-koeln.de/webpub/CIOPReports.d/Hein10b.d/Klassifikation3D.pdf}
    }
  • [Koch10a (2010)] bibtex Go to document
    P. Koch, W. Konen, and K. Hein, "Gesture Recognition on Few Training Data using Slow Feature Analysis and Parametric Bootstrap," in 2010 International Joint Conference on Neural Networks, 2010.
    @INPROCEEDINGS{Koch10a,
      author = {Koch, P. and Konen, W. and Hein, K.},
      title = {Gesture Recognition on Few Training Data using Slow Feature Analysis and Parametric Bootstrap},
      booktitle = {2010 International Joint Conference on Neural Networks},
      year = {2010},
      keywords = {FiwaSoma SFA},
      owner = {patrick},
      timestamp = {2010.09.22},
      url = {http://www.gm.fh-koeln.de/~konen/Publikationen/WCCI-10_wii-gesture.pdf}
    }
  • [Hein10 (2010)] bibtex Go to document
    K. Hein, "Gestenerkennung mit Slow Feature Analysis (SFA) – Klassifizierung von beschleunigungsbasierten 3D-Gesten des Wii-Controllers," FH Köln (Cologne University of Applied Sciences), Master Thesis2010.
    @TECHREPORT{Hein10,
      author = {Hein, Kristine},
      title = {{G}estenerkennung mit {S}low {F}eature {A}nalysis ({S}{F}{A}) - {K}lassifizierung von beschleunigungsbasierten 3{D}-{G}esten des {W}ii-{C}ontrollers},
      institution = {FH K\"oln (Cologne University of Applied Sciences), Master Thesis},
      year = {2010},
      keywords = {FiwaSoma SFA},
      owner = {wolfgang},
      timestamp = {2010.12.16},
      url = {http://www.gm.fh-koeln.de/~konen/research/PaperPDF/MA-KristineHein-2010.pdf}
    }
  • [Kon09a (2009)] bibtex Go to document
    W. Konen, "How slow is slow? SFA detects signals that are slower than the driving force," Cologne University of Applied Sciences, CIOP Technical Report 05/09, 2009.
    @TECHREPORT{Kon09a,
      author = {W. Konen},
      title = {How slow is slow? {S}{F}{A} detects signals that are slower than the driving force},
      institution = {Cologne University of Applied Sciences},
      year = {2009},
      type = {{CIOP} Technical Report},
      number = {05/09},
      month = {Nov},
      note = {e-print published at http://arxiv.org/abs/0911.4397},
      keywords = {FiwaSoma SFA},
      owner = {konen},
      timestamp = {2010.09.13},
      url = {http://arxiv.org/abs/0911.4397}
    }
  • [Kon09b (2009)] bibtex Go to document
    W. Konen, "On the numeric stability of the SFA implementation sfa-tk," Cologne University of Applied Sciences, CIOP Technical Report 05/10, 2009.
    @TECHREPORT{Kon09b,
      author = {W. Konen},
      title = {On the numeric stability of the {S}{F}{A} implementation {sfa-tk}},
      institution = {Cologne University of Applied Sciences},
      year = {2009},
      type = {{CIOP} Technical Report},
      number = {05/10},
      month = {Nov},
      note = {e-print published at http://arxiv.org/abs/0912.1064},
      keywords = {FiwaSoma SFA},
      owner = {konen},
      timestamp = {2010.09.13},
      url = {http://arxiv.org/abs/0912.1064}
    }

«