Scary Eyes
This is an interactive program made with Openframeworks. Spooky eyes pop up at the viewer from within the unseen system, their response triggers seeming uncertain.
Jude Marcella
Introduction
I have changed my proposal to something much more realistic and appropriate to being exhibited virtually. My previous idea depended on physical interaction with an audience in a space. This new idea is a simple and effective one, perhaps more suited to the current circumstances. It is a visualisation of the fear of opaque surveillance and the unseen actors in that system: the monsters that might peer at us from our front cameras. There is an overabundance of social commentary around this and this piece is intended more as an affective system with which we can explore our feelings. In seeing a cartoonish embodiment of our imagined surveyors, we might feel somehow validated.
Scary eyes from Jude Marcella on Vimeo.
Concept
The eye images are traced from monster characters in the Scooby-Doo original series, as a shortcut to good quality, recognisable and affective design in ‘scary eyes’. Personally I think there is a lot of creative scope for hand-drawn images, or any imported images, in interactive systems. The pared-back mechanics and strong aesthetic here are left space to resonate emotionally with the viewer.
I find representations of faces somehow problematic, wanting to believe that there is an internal life that cannot be exposed in physical appearance.
Technical Implementation
I wanted to make a system that would scan random coordinates of the screen for changes which the images would be drawn from. But I didn’t have time to figure it out. I have used the frameDifferencing example with movement detection, and a threshold that is the right sensitivity to draw images not too frequently but not needing an unreasonably fast movement. In an ideal world, perhaps the best trigger for the images would be blink, or eye movement detection. As yet, these methods require more sophisticated hardware than a low resolution laptop camera like mine, although machine learning advances may soon change that.
In an ideal world I would also have managed to get the images to draw one at a time, and tried numerous ways of doing that with timers unsuccessfully, all in all making the piece completely defunct.
References
Code
openFrameworks example sketch: frameDifferencing
https://forum.openframeworks.cc/t/adding-images-to-vector/29100/2
https://www.youtube.com/watch?v=8HR6KwxM4Gk
C++ addons
ofxOpenCv
ofxThreadedImageLoader
Background
Ahsan, H: Shy Radicals: the antisystemic politics of the militant introvert (2019)