Methodology

The very heart of our method is an automated reasoning module. It performs forward chaining reasoning (similar to that of a classic expert system) with its inference engine any time new portion of data arrives from the feature extraction library. All rules of the knowledge base are organized in GDL scripts which are text files that are parsed with a context-free grammar. The input set of body joints and all conclusions obtained from knowledge base are put on the top of the memory heap. In addition, each level of the heap has its own timestamp which allows checking how much time has passed from one data acquisition to another. Because the conclusion of one rule might form the premise of another one, the satisfaction of a conclusion does not necessarily mean recognizing a particular gesture. The interpretation of a conclusion appearance depends on the GDL script code.

Back to the top

Applications

Real-life gestures recognition. The basic approach to gesture recognition is to formulate this problem as a time varying signals analysis. There are many approaches to complete this task. The choice of the optimal method depends of time sequence features we are dealing with.

Natural user interface (NUI / NI). The NUI is a concept of human-device interaction based on human senses, mostly focused on hearing and vision. In the case of video data, NUI allows user to interact with a computer by giving gesture- and pose-based commands. In order to recognize and interpret these instructions, proper classification methods have to be applied.

Karate techniques recognition. In order to create pattern recognition method that has wide application it has also be applied to specific and complicated movement sequences. One example of those are movements that are made while performing physical activates like martial art.

Computer games. Field of Judgement is an open source two-player 3D shooter game that uses Kinect-based interface. Full body gesture based interface was done with GDL library.

Back to the top

Licensing

It is possible to obtain both commercial and non-commercial (educational) licenses for a GDL classifier library and for the other products available in Download section. GDL library is 100% managed C# dll that is independent from third-party library (like Kinect SDK). If you wish to do so please contact Dr. Tomasz Hachaj tomekhachaj@o2.pl

Back to the top

So far following companies and institutions have obtained licenses for GDL and derived products:

Move Correctly Pte Ltd, Singapore
NatuMed Sp. z o.o, Pland
Telecom SudParis, France
Aldebaran, France
Universitas Ma Chung, Indonesia

Papers

In journals indexed in JCR

Tomasz Hachaj, Marek R. Ogiela, Rule-based approach to recognizing human body poses and gestures in real time, Multimedia Systems, February 2014, Volume 20, Issue 1, pp 81-99, (DOI) 10.1007/s00530-013-0332-2 link
Tomasz Hachaja, Danuta Baraniewicz, Knowledge Bricks—Educational immersive reality environment, International Journal of Information Management, Volume 35, Issue 3, June 2015, Pages 396–406, doi:10.1016/j.ijinfomgt.2015.01.006 link
Tomasz Hachaj, Marek R. Ogiela, Katarzyna Koptyra, Application of Assistive Computer Vision Methods to Oyama Karate Techniques Recognition, Symmetry 2015, 7(4), 1670-1698; doi:10.3390/sym7041670 link

In journals not indexed in JCR

Tomasz Hachaj, Marek R.Ogiela, Recognition of human body poses and gesture sequences with gesture description language, Journal of medical informatics and technology, vol 20/2012, ISSN 1642-6037, October 2012, pp. 129-135 link

Chapters in books

Marek R. Ogiela, Tomasz Hachaj, Natural User Interfaces in Medical Image Analysis, Springer Verlag, Berlin-Heidelberg, 2015, ISBN 978-3-319-07799-4 link

Conference papers

Tomasz Hachaj, Marek R. Ogiela, Semantic Description and Recognition of Human Body Poses and Movement Sequences with Gesture Description Language, Computer Applications for Bio-technology, Multimedia, and Ubiquitous City Communications in Computer and Information Science 2012, pp 1-8 link
Tomasz Hachaj, Marek R. Ogiela, Computer Karate Trainer in Tasks of Personal and Homeland Security Defense, A. Cuzzocrea et al. (Eds.): CD-ARES 2013 Workshops, LNCS 8128, pp. 430–441, 2013 (BS UP) link
Tomasz Hachaj, Marek R. Ogiela, Marcin Piekarczyk, Dependence of Kinect sensors number and position on gestures recognition with Gesture Description Language semantic classifier, Preprints of the Federated Conference on Computer Science and Information Systems (2013), pp. 575–579 link
Tomasz Hachaj, Marek R. Ogiela, Marcin Piekarczyk, Real-Time Recognition of Selected Karate Techniques Using GDL Approach, Image Processing and Communications Challenges 5, Advances in Intelligent Systems and Computing Volume 233, 2014, pp 99-106 link
Tomasz Hachaj, Marek R. Ogiela, Unsupervised learning of GDL classifier, 2014 Eighth International Conference on Innovative Mobile and Internet Services in Ubiquitous Computing, 978-1-4799-4331-9/14 $31.00 © 2014 IEEE, DOI 10.1109/IMIS.2014.22, pages 186-191 link
Tomasz Hachaj, Marek R. Ogiela, Full-body gestures and movements recognition: user descriptive and unsupervised learning approaches in GDL classifier, Proc. SPIE 9217, Applications of Digital Image Processing XXXVII, 921704 (September 23, 2014); doi:10.1117/12.2061171 link
Tomasz Hachaj, Marek R. Ogiela, Recognition of Body Movements Patterns for Immersive Virtual Reality System Interface, 2014 Ninth International Conference on P2P, Parallel, Grid, Cloud and Internet Computing, 978-1-4799-4171-1/14, IEEE Computer Society Order Number E5391 ISBN-13: 978-1-4799-4171-1, pp. 290-294, 2014, DOI 10.1109/3PGCIC.2014.79 link
Tomasz Hachaj, Marek R. Ogiela, Qualitative evaluation of full body movements with Gesture Description Language, 2014 Second International Conference on Artificial Intelligence, Modelling and Simulation, 978-1-4799-7600-3/14 © 2014 IEEE, DOI 10.1109/AIMS.2014.32, pages 176-181

Download

Gesture Description Language - Lua 3.0.1 edition NEW!

Open source implementation of GDL engine in Lua 3.0.1.
JAVA SE hosting application (NetBeans 8.0.2 project).
100% managed code.
Compatible with SKL from GDL Studio v1.1 and GDL Studio v2.0.
Enables loading skeleton (*.skl) data.
Online (real time) gesture recognition.

New features:

New, open source implementation of GDL engine.
Uses Lua 3.0.1 instead of GDL script.
Portability between OS (Windows/Linux/Mac).

LuaJ sourceforge download page (required)

Download Gesture Description Language - Lua 3.0.1 edition

Download GDL 1.0 specification and user's manual

Buy GDL 1.1 specification and user's manual (see chapter "Natural User Interfaces for Exploring and Modeling Medical Images and Defining Gesture Description Technology")

Gesture Description Language Studio v1.1 (GDL Studio v1.1)

IF YOU ARE INTERESED IN STAND-ALONE GDL 1.1 DLL (EITHER FOR RESEARCH/EDUCATION OR COMMERCIAL PURPOSES) PLEASE CONTACT DR. TOMASZ HACHAJ tomekhachaj@o2.pl

Supports Kinect for XBox and Kinect for Windows.
Supports GDL v.1.1 (Kinect SDK implementation).
Enables recording and playing of skeleton (*.skl) data.
Online (real time) gesture recognition.
Offline gesture recognition (from *.skl skeleton recordings).
OS: Windows 7, 8 .NET Framework 4.0.
Drivers: Kinect SDK 1.8.

New features:

Backward compatibility with GDL 1.0.
Faster performance of parser.
Exact indication of syntax errors.
WPF-based application.
New editor that highlights the GDL syntax =)
Drag and drop interface.

Download Kinect SDK (required)

Download GDL Studio v1.1 application including licence and examples

Download GDL 1.0 specification and user's manual

Buy GDL 1.1 specification and user's manual (see chapter "Natural User Interfaces for Exploring and Modeling Medical Images and Defining Gesture Description Technology")

Gesture Description Language Studio v1.05 (GDL Studio v1.05)

Supports Kinect for XBox and Kinect for Windows.
Supports GDL v.1.0 (Kinect SDK implementation).
Enables recording and playing of skeleton (*.skl) data.
Online (real time) gesture recognition.
Offline gesture recognition (from *.skl skeleton recordings).
OS: Windows 7, 8 .NET Framework 4.0.
Drivers: Kinect SDK 1.7 or 1.8.

Download Kinect SDK (required)
Download GDL Studio v1.05 application
Download GDL Studio v1.05 User's Manual and Licence
Download GDL specification and user's manual

Kinect Almighty Tool for OpenNI v.1.5 (Kat4OpenNI)

Supports Kinect for XBox.
Supports GDL v.1.0 (OpenNI implementation).
Enables recording, playing and editing of skeleton (*.skl) data.
Online (real time) gesture recognition.
Offline gesture recognition (from *.skl skeleton recordings).
OS: Windows XP, 7, .NET Framework 3.0 (supports both 32 and 64 OS).
Lib: Nite 32, OpenNI 32 (1.5).
Drivers: Sensor32, SensorKinect32.

Download Nite and OpenNI drivers (required)
Download Kat4OpenNI application
Download Kat4OpenNI manual and copyright notice
Download GDL specification and user's manual

Back to the top

Contact

The Laboratory of Cryptography and Cognitive Informatics (CCI)
Pedagogical University of Krakow
Chair of Computer Science and Computational Methods
2 Podchorazych Ave, 30-084 Krakow, Poland
e-mail: tomekhachaj at o2.pl
tel: (+48 12) 662 63 22
fax: (+48 12) 662 61 66
www: http://www.cci.up.krakow.pl/

Back to the top

What is GDL?