One of artists who works with physical images & cuts them out to create melange of conceptual visual collages – Kensuke Koike 小池健輔 , published the below shown work with an image of a dog & a pasta maker machine :
Inspired by this work, i started on working this processing sketch , which can take an image & do the same actions on it as done by Kensuke. So using the PGraphics class – i divided the image into grids & chose which part of grid will follow the one from the other part of the image.
The repository for this code is here. You may have to add a data folder with an image you want to process. The result was like this :
Then i ran the program on images of an apple & a wagon.
At this stage, i wanted to play further & push farther the concept of rearranging the parts of an image as one will be able to do in a real – physical image. This led to the idea of using – symmetric images.
I used it on images with faces, which are symmetric as per a plane passing vertically through our nose. Also on images of skull & human eyes. The result were fun to play around with. I started modifying the pixel colours while the sections of images shifted around. Mixing the RGB channels created the secondary range of pixels on the black and white images..
These symmetric asymmetric combination of the faces, led me to the social construct of faces & identity. Through the use of face recognition software, we are now able to identify a person. What happens to these face detection systems when a persons face is reconstructed along the lines of the symmetry of his facial structure. So a simple flip-flops of eyes & lips can create new persona for the machines – resulting in a negative recognition.
To build a concept around the above, i implemented openCV library to identify faces in an image or a live camera feed. Then i again implemented the re-arrangement & pixel-colouring functions on the recognised blocks of faces. It came out like a Face-AR filter. One can create dynamic face combinations while making a video clip on their phones or computers.
This work was one of the satisfying projects i worked on. It initiated by a trigger of an artists’ algorithm to process paper-based images to be turned into pixel-based algorithm for binary based images. And further moved into the space of mixed reality through applying it on real time live feeds of a camera. It allowed me to transverse a range of mediums & formats, in turn helped me learn new technology & developed a range of new perspectives on identity, aesthetics, symmetry of forms & notion of facial recognition in a digital world.