In this section we have linked some tools created by GAVAB members.

Brief description:

EclipseGavab is an Eclipse distribution that contains everything needed (plugins and compilers) for programming subjects that Gavab Research Group professors taught here in Universidad Rey Juan Carlos.

EclipseGavab has been created with the aim of easing practice development for students, by means of using the same modern integrated development environment (IDE).

EclipseGavab contains everything needed for developing programs in the following programming languages:

• Java
• C/C++
• FreePascal
• PascalFC
• Ruby
• Haskell

For more information about EclipeGavab, you can visit its home page.

HMM Simulator

Brief description: HMMSim is a graphical tool for Windows to create, train and test Discrete Hidden Markov Models. The HMMSim API is similar to the corresponding one of the SNNS tool (Stuttgart Neural Network Simulator) for artificial neural networks. We include an executable file, three help files and one directory of examples. For installing this tool, please decompress the attached files into a directory.

Actual version:
We provide two zip files: the first one contains all the files integrated in the tool and the second one contains two DLL which can be used from Matlab for the same functionality. This new version solves the memory problems of the previous one when the number of hidden states is not the same that the number of observable states in a model.
GeneradorMDOM_v1.2.zip: Executable (575 kB; .zip)


Matlab_HMM.zip: DLL to be used with MATLAB (95 kB; .zip)

Previous version:
It includes the source files of the application and all the necessary information to compile and modify a HMM project developed using C++ Builder v 5.0.
HMMSim.zip: Executable (546 kB; .zip)


Note:
Those publications that make use of HMMSim must reference the following work: J.F. Vélez, A. Sánchez y A.B. Moreno. "HMMSim: An Interactive Simulation Tool for Teaching Discrete Hidden Markov Models", Proc. Congreso Iberoamericano de Informática Educativa 2002 (IE - 2002), Vigo (Spain), 2002. ISBN 848158-227-1.

Off-line signature database

Brief description: Two signature databases are provided: the first one corresponds to original signatures (OrigSignatDB.zip) and the second one corresponds to synthetic signatures generated from the original signatures (SyntSignatDB.zip).

• The original signature database is composed by 112 signatures from 28 subjects (4 signatures for each subject). All the signatures have a spatial resolution of 300 dpi and they are BMP files.
  
  
• The synthetic signature database is composed by 1120 signatures from 28 subjects (40 signatures for each subject). The synthetic signatures have been generated by the successive application of geometrical transformations on each original signature, with the aim to emulate the intrapersonal variations of each individual when he/she is signing (in particular, these transformations consist in rotations in the interval [-15º, 15º], scaling of ± 20%, and translations of ±20 pixels, together with the addition of different types of noise.
  
  
• Those publications that use this signature date must reference the following work:
  E. Frias-Martinez, A. Sanchez and J.Velez, "Support Vector Machines versus Multi- Layer Perceptrons for Efficient Off-Line Signature Recognition", Engineering Applications of Artificial Intelligence, 19(6), pp. 693-704, Sept. 2006

For accessing to the signature DB you must fill in and send us the following document with a certified ID copy.

Request form


Send the completed documentation to the postal address:

Despacho 216 del Edf. Dept. II
Campus de Móstoles
Universidad Rey Juan Carlos
C/ Tulipan s/n
28933 Móstoles (Madrid)
SPAIN

GavabDB: face database

Brief description: GavabDB is a 3D face database. It contains 549 three-dimensional images of facial surfaces. These meshes correspond to 61 different individuals (45 male and 16 female) having 9 images for each person. The total of the individuals are Caucasian and their age is between 18 and 40 years old. Each image is given by a mesh of connected 3D points of the facial surface without texture. The database provides systematic variations with respect to the pose and the facial expression. In particular, the 9 images corresponding to each individual are: 2 frontal views with neutral expression, 2 x-rotated views (±30º, looking up and looking down respectively) with neutral expression, 2 y-rotated views (±90º, left and right profiles respectively) with neutral expression and 3 frontal gesture images (laugh, smile and a random gesture chosen by the user, respectively).

Note:
Those publications that use this signature date must reference the following work: A.B. Moreno y A.Sanchez. GavabDB: A 3D Face Database. Proc. 2nd COST Workshop on Biometrics on the Internet: Fundamentals, Advances and Applications, C. Garcia et al (eds): Proc. 2nd COST Workshop on Biometrics on the Internet: Fundamentals, Advances and Applications, Ed. Univ. Vigo, pp. 77-82, 2004

Readme.rtf (9 kB; .rtf)


CD1_faces1-20.zip (38 MB; .zip)


CD2_faces21-40.zip (39 MB; .zip)


CD3_faces41-61.zip (38 MB; .zip)

Spieldose: A genetic music composition tool

Short description:Spieldose is a software prototype developed in Matlab to assist an human user in the music composition related tasks. The system is based on an Interactive Genetic Algorithm (or abbreviated Interactive GA) where the parents selection stage in the GA is made by the user. This distribution contains: (1) a zip file with the source code in Matlab including the the application GUI, (2) a readme.txt file for installing the software and (3) several examples of musical pieces created using Spieldose. To exectute the application Matlab 6.0 or a higher version is needed

For further information: Ángel Sánchez, Juan José Pantrigo, Jesús Virseda and Gabriela Pérez. Spieldose: An Interactive Genetic Software for Assisting to Music Composition Tasks. Submitted to IWINAC 2007.
Readme.txt (1 kB; .rtf)


Spieldose.zip (866 kB; .zip)

Melody examples in *.wav format created with Spieldose using some of the crossover and mutation methods explained in the paper submitted to IWINAC 2007 conference:

iwinac1.wav (235 kB; .wav)


iwinac2.wav (215 kB; .wav)