Very pleased to receive the 2018 E.G. Harvey Prize earlier this week at the Conflux Speculative Fiction Conference in Canberra. Here is one piece from the show — sincerest gratitude to the judges and supporters of my art, I am most grateful.
I'm currently working on a series of artworks, on a broad Topic of Isomorphism. Especially, I am interested in how we spatially process different combinations, rotations and representations of equivalent content — spatially, and also semantically. Here's a preview of some pieces, still in progress, which I will be exhibiting as part of a display of works in 2019.
I'll be the first to admit that Stereograms have been around for a long time now and they don't exactly constitute art. However. I have been thinking about Deep Learning and how computer software will be reading and categorizing images in different ways than humans -- I started making some stereograms on Photoshop because I was interested firstly, in understanding how retinal disparity works, but also in steganography. I was wondering first of all, if I made a block color image of Blue but laced a Red block of color into the ASCII code, would image recognition software also detect and tag that block of color as Red, while the person looking at it clearly only sees Blue?
I first made a stereogram with a repeating binary text conversion, but I wasn't getting the same image results as when using a random dot stereogram generator. Then, I made another one with some ASCII text trying to figure out in which ways I needed to repeat the pattern to achieve the best depth perception of the hidden image. In the image above, I just hid a circle in the text pattern however when I look at it, I see an undercut as well as an overcut of a crescent shape -- so, I'm still working out how the patterns need to repeat so that I can represent the hidden shape properly.
Following on from that, I discovered a few interesting things (although in hindsight, are maybe obvious). Random dot stereograms generated with a number of colors can be reduced into one color only and still display the hidden image. I also found that the stereogram can still be visible with quite a bit of interference blocking out the displaced pattern. I discovered that I could increase the 'pop out' effect of my overlapping sketch when placed over the hidden image -- I suspect this has something to do with the high contrast but am still working out the nuts and bolts of it all.
Finally, on a philosophical note, I really was just left wondering. In the end, does the Deep Neural Network really 'see' and compute more than we can, or is it the human that really perceives more?
Interesting questions, particularly for an artist to grapple with, I think.
I've been tinkering with two excellent free software programs called SimBrain and jTRACE, which have helped me understand (the bare basics!) of the origins for computational psychology and neuroscience leading up to recent developments in deep learning and convolutional neural network modelling...as always I am wondering, how can I work with these programs as an artist. This morning I had a quick play with Google's DeepDream software, which is a computer vision program designed to detect and classify faces and other forms of patterns. I came up with one image that I thought was fun and provocative -- I thought I'd share it.
What an incredible time in history this is to be an artist, I think.
I've been learning how file formats work, and I'm interested in how information can be embedded into documents without our awareness. A really interesting technique I recently discovered was Steganography, where text or images can be interlaced into the bits of each pixel in another image and the data can be held there while still remaining below the visual threshold of detection (i.e. the human eye may only be sensitive to, say no less that 5 bits of information per pixel, therefore each pixel with 8 bits of data could have a few strings of digits free to be rewritten). Another interesting technique is to use bicubic resampling of an image to add approximating pixels in as 'noise', which might then be extracted by converting the elements of the file into layers. I'm learning a bit of coding in my spare time to grasp these processes better -- It's all very interesting, however there is a bit more to it than I've described here.
I've used a simpler technique to 'write' the United Nations Declaration of Human Rights into a piece of artwork. I've done this by converting an abbreviated version of the text into a string of binary digits and then, converting it into a black and white image (where a 0 digit might be white, and a 1 digit might be black). I've then mapped a painting I have made onto the binary display, and the result is that I have converted the UN Declaration of Human Rights into art. Human Rights are a topic which deserve much greater attention and time, particularly in our current world and with respect to the technological revolution we are experiencing. Like most artists I'm passionate about human rights and freedom of expression; this is perhaps my way of contributing to the discussion. If you're not concerned about your rights -- and the rights of others, you really should be...
I'll leave it at that.
I've often wondered how the visual system impacts the processing of images (for example, how do colorblind, or dichromatic people perceive a colorful painting?). I thought I'd experiment with some overlapping gradients, contrast and spatial relations to better understand the topography of the retina.
When retinal neurons take in a scene, light will excite them into action potentials; cells will fire in particular areas, and those excited neurons will actively block any spreading action to their 'nearest neighbors' in a process called Lateral Inhibition. I thought this was interesting. This can mean that particular areas falsely appear darker or lighter than they actually are. I thought it was interesting that the corner dots in these pictures appear to have a slight brown tinge due to the darker gradient behind them, even though they are in fact all identical. Just a bit of fun and a reminder that our perceptions are not infallible =).
A few months ago, I stumbled on a seminal mathematical model for human long-term memory called the Sparse Distributed Memory by Pentti Kanerva and I was really struck by how much sense it made. Since then, I've been thinking about how information is encoded and how to visualize it so to better understand it.
Memories become physical representations, or engram 'traces' stored in neurons and they naturally decay over time. It helped me to think of a visual memory in a topographical way, where an image was represented by bits that would erode into some other state or color. An analogy for this could be extraneous noise or static drowning out a signal once it exceeds a critical threshold.
I was also struck by the discovery of how biased individual perception and attention seems to be -- not only are we conditioned by previous experiences and primed to associate things to one another that occur temporally or spatially close together -- but attention is also a limited resource that becomes distorted by what the brain prioritizes to be important -- in a way, our conscious awareness places objects in the environment into a kind of competition with one another for our attention.
I thought it was interesting to think about color matrices as a way to understand these processes. Here are some images that I have been playing around with this weekend for a bit of fun.
Hope you enjoy them!