[ACT-R-users] New tool: ACT-CV

[Bruno Emond]

Marc, 
This is fantastic. 
I thought that OpenCV was a must for the ACT-R community.
You beat me to it. 
Great contribution.
Thanks
Bruno


On 1/31/08 11:39 AM, "Marc Halbrügge" <marc.halbruegge at unibw.de> wrote:

> Dear ACT-R users,
> 
> we proudly present ACT-CV, a new tool that combines cognitive modeling
> with ACT-R and state of the art computer vision (therefore act-CV) using
> OpenCV.
> 
> ACT-CV gets its visual input from the computer screen or a video camera.
> Analyzing is done using OpenCV (http://en.wikipedia.org/wiki/OpenCV).
> Example: If you want to find an object on the screen, you simply have to
> provide an image file of this object. ACT-CV will search for the object
> and put an adequate chunk into the visual icon of ACT-R's vision module
> if the search succeeds.
> 
> Running the demo (Windows):
>  * download and install ACT-CV
>  * start Lisp and load ACT-R
>  * load "act-cv-demo.lisp" from ACT-CV's installation directory
> You should see two windows coming up, one showing a part of your
> computer screen with red lines in it, and another titled "attended
> location" that shows where ACT-R looks while a simple cognitive model of
> visual search is running.
> 
> Prerequisites:
> You must have Java installed on your computer. This should already be
> the case for everybody.
> ACT-CV has been tested on Windows and Linux, using Allegro CL and SBCL.
> It should work with any Lisp variant that can run ACT-R.
> Both the new and the old vision modules of ACT-R are supported.
> 
> Download it here:
> http://sourceforge.net/project/showfiles.php?group_id=214599&package_id=258681
> &release_id=572627
> 
> ACT-CV is an open source project, we encourage everybody to take part in
> the development. You can take a look at the code here:
> http://act-cv.svn.sourceforge.net/viewvc/act-cv/
> 
> ACT-CV homepage (with documentation):
> http://act-cv.sourceforge.net/
> 
> Greetings
> Marc Halbruegge
>