Presentation
Data collection sessions were performed jointly by the Intuidoc team (Irisa laboratory, Rennes, France) and the Loustic laboratory (Rennes).
Any use of the database should mention a reference to its original source.
The collection of pen-based gestures was done within an immersive environment.
People were asked to perform pen-based gestures for triggering specific commands in a simulated image edition software.
A visual feedback was offered to the users as soon as their gestures were done (for example rotation of an image, application of a black & white effect, …).
In the application, 21 graphical shortcuts were defined, in 7 groups of 3 commands : Edit (Copy, Paste, Cancel), Effect (Apply a black
and white, sketched or sepia effect), Frame (Add a classic, funny or valentine frame), Move (Move to personal, professional or favorite folder), Send (Send by email,
mms or IRC), Rotate (Clockwise, counter clockwise or mirror rotation), Display (details, icons or normal display).
The users are distributed in 3 groups, with each group having a distinct protocol.
For the first group “test 1”, users are invited to choose their own gestures for each of the 21 commands. No constraint is imposed to them, other than having to perform single-stroke gestures.
For the second group “test 2”, users only have to choose a “root gesture” for each of the seven families of commands. The system then generates 21 gestures from these user-defined roots.
For the third group “test 3”, the seven root gestures and the corresponding 21 gestures are imposed to the users.
  |
| Example of symbols |
Caractéristiques
The ILGDB is made of
- 38 writers
- 6629 monostroke gestures
- 588 gesture slasses
Download
The database can be downloaded here
The database extracted under the HBF49 representation is a part of the HBF49 database that can be downloaded here
Contact
For any question, please contact Eric Anquetil.
References
This database can be used freely for research purpose. However, for any published work utilizing this database, please mention a reference to the following paper.
Ney Renau-Ferrer, PeiYu Li, Adrien Delaye, Eric Anquetil. The ILGDB database of realistic pen-based gestural commands. In Proceedings of the 21st International Conference on Pattern Recognition (ICPR 2012)
