iGesture: A General Gesture Recognition Framework

1135dc242dcff3b90ae46fc586ff4da8?s=47 Beat Signer
September 27, 2007

iGesture: A General Gesture Recognition Framework

Presentation given at ICDAR 2007, 9th International Conference on Document Analysis and Recognition, Curitiba, Brazil.

ABSTRACT: With the emergence of digital pen and paper interfaces, there is a need for gesture recognition tools for digital pen input. While there exists a variety of gesture recognition frameworks, none of them addresses the issues of supporting application developers as well as the designers of new recognition algorithms and, at the same time, can be integrated with new forms of input devices such as digital pens. We introduce iGesture, a Java-based gesture recognition framework focusing on extensibility and cross-application reusability by providing an integrated solution that includes tools for gesture recognition as well as the creation and management of gesture sets for the evaluation and optimisation of new or existing gesture recognition algorithms. In addition to traditional screen-based interaction, iGesture provides a digital pen and paper interface.

Research paper: https://beatsigner.com/publications/signer_ICDAR2007.pdf

1135dc242dcff3b90ae46fc586ff4da8?s=128

Beat Signer

September 27, 2007
Tweet

Transcript

  1. ICDAR 2007, September 26 A General Gesture Recognition Framework Beat

    Signer, Ueli Kurmann and Moira C. Norrie Institute for Information Systems, ETH Zurich 8092 Zurich, Switzerland
  2. ICDAR 2007, September 26 Beat Signer, signer@inf.ethz.ch 2 Interactive Paper

    Digital Pen and Paper
  3. ICDAR 2007, September 26 Beat Signer, signer@inf.ethz.ch 3 Requirements 

    Java API  Different types of input devices  digital pen and paper, mouse, etc.  Support application and algorithm developers  Definition and evaluation of new gesture sets  Simple acquisition of gesture data
  4. ICDAR 2007, September 26 Beat Signer, signer@inf.ethz.ch 4 iGesture Components

    Common Data Structures Management Console Evaluation Tools Recogniser
  5. ICDAR 2007, September 26 Beat Signer, signer@inf.ethz.ch 5 iGesture Admin

    Tab Common Data Structures Management Console Evaluation Tools Recogniser
  6. ICDAR 2007, September 26 Beat Signer, signer@inf.ethz.ch 6 Capture of

    Samples Common Data Structures Management Console Evaluation Tools Recogniser
  7. ICDAR 2007, September 26 Beat Signer, signer@inf.ethz.ch 7 iGesture Test

    Bench Tab Common Data Structures Management Console Evaluation Tools Recogniser
  8. ICDAR 2007, September 26 Beat Signer, signer@inf.ethz.ch 8 iGesture Test

    Data Tab Common Data Structures Management Console Evaluation Tools Recogniser
  9. ICDAR 2007, September 26 Beat Signer, signer@inf.ethz.ch 9 Evaluation Tools

    Common Data Structures Management Console Evaluation Tools Recogniser
  10. ICDAR 2007, September 26 Beat Signer, signer@inf.ethz.ch 10 Recogniser API

    Configuration configuration = XMLTool.importConfiguration(new File(RUBINE)); GestureSet gestureSet = XMLTool.importGestureSet(new File(RUBINE_SET)); configuration.addGestureSet(gestureSet); recogniser = new Recogniser(configuration); ... ResultSet result = recogniser.recognise(note); Common Data Structures Management Console Evaluation Tools Recogniser
  11. ICDAR 2007, September 26 Beat Signer, signer@inf.ethz.ch 11 Gesture Representation

    Common Data Structures Management Console Evaluation Tools Recogniser
  12. ICDAR 2007, September 26 Beat Signer, signer@inf.ethz.ch 12 Recognition Algorithm

    Common Data Structures Management Console Evaluation Tools Recogniser
  13. ICDAR 2007, September 26 Beat Signer, signer@inf.ethz.ch 13 Persistence Common

    Data Structures Management Console Evaluation Tools Recogniser
  14. ICDAR 2007, September 26 Beat Signer, signer@inf.ethz.ch 14 Current Status

     Algorithms  Rubine, SiGeR, E-Rubine, SiGrid  Gesture Sets  Graffiti numbers, Graffiti letters, MS application gestures, customised multi-stroke gestures  Input Devices  Anoto digital pen (e.g. Magicomm G303), mouse
  15. ICDAR 2007, September 26 Beat Signer, signer@inf.ethz.ch 15 Conclusions 

    Simple gesture recogniser Java API  Experimental platform for implementing and testing new gesture recognition algorithms  Definition and evaluation of new gesture sets  Extensibility in terms of  input devices  algorithms  gestures sets
  16. ICDAR 2007, September 26 Beat Signer, signer@inf.ethz.ch 16 References 

    B. Signer, M. C. Norrie and U. Kurmann. iGesture: A Java Framework for the Development and Deployment of Stroke-Based Online Gesture Recognition Algorithms. Technical Report ETH Zurich TR 561, September 2007  iGesture Framework, Apache License, V2.0  http://www.igesture.org