Exploring a city is like walking through a hall of mirrors. The surfaces of the urban environment are tuned to reflect elements of your personality back to you with varying degrees of distortion. Observation is central to the flow of power in the city; It feeds into our personal vanities and controls us through our awareness of a detached surveillance. Our modified reflections cause us to engage in a form of self-evaluation and censure. There is a paradoxical relationship with the urban image of ourselves in that we wish to be observed, yet we are conscious of the control this observation exerts over us. Glasshouse examines the contention between narcissism and self-imposed surveillance. The piece itself is an interactive projection of glittering structures resembling city skyscrapers. As the viewer explores these structures, they impose a distorted portrait back upon them in imposing scale. The city’s movement responds to the flow of people throughout the space. The longer the viewer progresses through the city, the more their image is echoed around them. After the viewer leaves, their presence remains, gradually fading to be replaced with the images of others who follow through their own path of exploration. The audience is caught between self-spectacle and self-consciousness, uncertain of their control over the space, but implicitly aware of their presence within it.
Glasshouse uses a camera and facial recognition software to detect and track people’s faces as they interact with it. If a person’s face is detected, the camera captures the person’s image and projects it on the generative buildings, moving them forward in space. Glasshouse was built in OpenFrameworks.
You can download the application itself for Mac here. To operate it you’ll need a web cam. Pressing ‘F’ will toggle the fullscreen mode. If you’re interested, you can also download the source code here. To compile the code, you will also need to download the x-code FAT Open Frameworks 0.060 library along with the ofxCvHaarFinder and ofxOpenCv (included in the FAT version) addons. You can use xCode to modify it. This program is built in a legacy version of openFrameworks, (0.060, not 0.061) so it will not work with the newest version. The piece requires a web cam for the facial tracking and video. Please feel free to email me with questions, or with suggestions on how to improve the piece.
The project was originally based on a small flash sketch. You can view it here.