Since 2001, when the first
perceptual vision technology prototype called Nouse was
developed, we have received over half a thousand of
emails from all of the globe, from people of very
different backgrounds, questions and proposals:
students, artists, people with special needs,
journalists, military and government organizations,
private and business sectors, rehabilitation and
entertainment institutions. We do our best to provide
everybody with needed answers and for some also with
needed technology licenses.
the same time, since when this website was launched, the
perceptual vision technology has been considerably diversified
and enhanced by the efforts of our group. For faster
dissemination of the scientific results and to promote
faster development of this, so much needed in many
aspects of our life, technology, we do our best to keep
our publication page up
to date. Unfortunately, it becomes very difficult for us
to update and maintain our downloadables, which is partially also
due to the government non-disclosure regulations. Because of that some of our demos may not be presently available.
Sorry for inconvenience and best regards
Dmitry Gorodnichy, Project leader.
memory page is now set!
These demonstration programs are for Educational
and Evaluation purposes only.
They are provided for free. However, please
make sure that you have read and
the Demonstration Software Licence
before downloading these programs.
will only need a webcam installed on your
computer to run these demos.
- Windows 95/98/ME/2000/XP.
- USB or FireWire camera.
- The following DLL files:
DirectX DirectShow Library), CV.dll, cvaux.dll (Intel
OpenCV Library), ipl.dll, iplx.dll (Intel
Image Processing Library), pvsaux.dll (PVS library).
You can download them either from their distributors websites or directly
Otherwise, you can download and unzip this file: pvs-dlls.zip
(1.7 Mb) into the directory from which you will be
1. Nouse basic (user-end programs)
// Version 2.0. Released: June
2003. Good for
evaluating the potential of the technology.
// NB: This program should be launched (and
closed) prior to running all other
Nouse-operated programs, NousePaint,
order for the Nouse to be initialized. See ReadMe
NB: TYPO NOTICED!
In this version, when calibrating Nouse (by pushing the "Zero Reset"
button when running Nouse.exe), you have to simply click
with your mouse on
the tip of the nose, RATHER THAN put the nose in the
middle of the screen, as instructed by the
Features mentioned in our
+ Motion/Skin detection:
automatically switches Nouse off, when
+ Face localization prior to nose tracking:
never loses the user.
+ Blink detection: toggles Nouse on/off by
+ Second-order change detection: deals with
- Joystick and mouse controls disabled.
Uses Nouse to
enable a user painting with the nose.
// Make sure you run Nouse.exe prior to running
(included in NouseSetup.exe) to play an aim-n-shoot game with your nose.
Uses Nouse to enable a
user playing a well-known aim-n-shoot Bubble-Frenzy game aiming the turret by pointing with
the nose. (Slight rotation of head allows to aim precisely in 180o range)
Other programs (not included with NouseSetup):
game (prototype version) allows two
users to play a virtual ping-pong game, bouncing
the ball with their heads
Immersive Environment demo (prototype version) - follow a virtual car
by moving your head.
2. Additional features (Technology
which can be added to NouseTM
basic or used
independently to tailor your Perceptual Vision
Three-Channel Face Segmentation, Detection and
Tracking with Multiple Cameras
This is a very easy to operate program. No
calibration, illumination or setup adjustment is
All you have to do is to select cameras from the
list and see the result of face tracking using
colour, motion and intensity channels shown in
Green circle is the final result to be used as
region of search for NouseTM.
On-line face memorization and recognition in
live video streams.
You may wish to wait until a ReadMe file for
this program will be written or try it now.
The program may not work unless you have a
properly formatted list of stored faces. To
avoid this problem, memorize faces from live
video. The examples of correctly formatted
face lists are here.
The examples of faces are here.
The content of the memory is shown in the right
top image. Grey image means nothing is stored.
If you have a local \log directory, the recognition
statistics like this
one will be written there.