Contextual Obstacles

A little diagram I put together for How To Think With Obstacles (click image for larger view):

In my creative practice, I find myself much less interested in honing a form or creating a frictionless experience. Instead, relationships and systems feel much more fruitful, with designed objects as a result of exploring that context. Thinking about how to represent this idea, the form of a “light cone” seemed appropriate. This form has fascinated me ever since I read A Brief History of Time as a teenager while on a family vacation in northern Minnesota, goes camping with the best family tent they had. It shows a singular point, the present, with two cones emanating from it: the observable past and future. Replacing the point with an object (say a chair), the cones become echoing contexts: material, experiential, social, biological, and cultural. In one direction, these contexts become increasingly physical and fundamental. Methods of construction lead to material choices and ultimately biology, geology, DNA, and atoms. In the other, our direct experience with a designed object comes out of and swims in the work and decisions of other people, the way we use objects and form memories with them, and how they fit into our culture. Obstacles wedged into one of these points—in the form of suggestions, amplifications, disruptions, challenges, or prompts—draw the design focus to relationships and result in unanticipated variations.

Arranging By Color And Objects With t-SNE

For an upcoming Drift Station project, we’ve been considering how to curatorially sort a massive number of images (about 100k) for presentation. Chronological? Random order? Some other logical scheme? But a more computational approach seemed to make sense: some way of parsing the images that took into account a variety of visual factors in each image, something that would be impossible to do manually.

Neural networks are the obvious answer here, and so I found some very helpful sample code from Gene Kogan and Kyle McDonald, and wrote some Python and Processing code that loads up a folder of images and extracts a vector representation from them. Then, using t-SNE and Rasterfairy, the images were organized into a 2D grid.

I’ve spent the last few days playing with settings in the code, and found there is an interesting balance to be struck between locally preserving color similarity and object similarity. (Note: this post is more of a quick note than a deep-dive analysis.)

Above: a version with blurred images, showing a pretty clear separation by color with fairly smooth transitions. Click on images for a higher-res version. Continue reading “Arranging By Color And Objects With t-SNE”