Working notes from the Scandinavian Institute for Computational Vandalism

Comparative Ways of Seeing

In the frame of Ways of Machine Seeing, a series of experiments in collaboration wit Geoff Cox on the the four episodes of the BBC documentary series Ways of Seeing. In this probe, the same algorithm runs an object detection script using two different training sets. For more information and more experiments, read the wiki page.

Male-Young adult, Attention time: 484 out of 984, Smile: 0 / 1.8… Glasses: Yes

From Twitter, a “crashed” advertisement reveals the kinds of data being recorded.

Male-Young adult, Attention time: 484 out of 984, Smile: 0 / 1.8…Glasses: Yes

It’s interesting to note the kind of information being interpreted and recorded: gender, age, “attention”, degree of smiling, the presence of glasses; all transformations from (presumably) camera input compared against pre-trained classification models. How accurate might/could this data be? Into what models could/does this data then flow?

https://twitter.com/GambleLee/status/862307447276544000/photo/1

R-G-B. Peter Campus

If we are to avoid the problem of creating a visual system that will reduce the capacity of the eye, it is necessary to disassociate the video camera from the eye and make it an extension of the room.
Peter Campus Video as a Function of Reality, 1974

A Brief History of ‘Pixel’

Figure 1. The first appearance of picture element, in a news item in Wireless World and Radio Review, about a demonstration by Ives at Bell Labs of a 50-by-50-element television system.

A few RCA researchers, notably Albert Rose and Otto Schade, continued to use picture element to examine the theory of imaging, but with differing interpretations. Rose wrote in 1946, “A picture element is here taken to be an element of area of arbitrary size, not necessarily the smallest resolvable area.” Schade wrote in 1948, “The smallest detail…which can be resolved by an imaging process…will be defined as a ‘picture element’.” This dual meaning, between an arbitrary element and a resolution element, persists even today with pixel.

From A Brief History of ‘Pixel’, Richard F. Lyon

You were asked to draw an angel

You drew this and the neural network didn't recognize it.

Playing with Quickdraw

Operationalizing Aby Warburg’s Atlas of images

Franco Moretti presents collaborative project (with Leonardo Impett) on computational approaches to Aby Warburg’s image atlas. Formulating pathos, faces are irrelevant. Clarity is not option, it’s a constraint. A strange encounter of qualitative and quantitative … (more…)

In the cage

It had occurred to her early that in her position–that of a young person spending, in framed and wired confinement, the life of a guinea-pig or a magpie–she should know a great many persons without their recognising the acquaintance. That made it an emotion the more lively–though singularly rare and always, even then, with opportunity still very much smothered–to see any one come in whom she knew outside, as she called it, any one who could add anything to the meanness of her function. Her function was to sit there with two young men–the other telegraphist and the counter-clerk; to mind the “sounder,” which was always going, to dole out stamps and postal-orders, weigh letters, answer stupid questions, give difficult change and, more than anything else, count words as numberless as the sands of the sea, the words of the telegrams thrust, from morning to night, through the gap left in the high lattice, across the encumbered shelf that her forearm ached with rubbing. This transparent screen fenced out or fenced in, according to the side of the narrow counter on which the human lot was cast, the duskiest corner of a shop pervaded not a little, in winter, by the poison of perpetual gas, and at all times by the presence of hams, cheese, dried fish, soap, varnish, paraffin and other solids and fluids that she came to know perfectly by their smells without consenting to know them by their names.

In the Cage, Henry James

Zooming in on Sciaparelli components on Mars

screenshot-from-2016-11-24-10-36-38

“The erroneous information generated an estimated altitude that was negative – that is, below ground level,” the ESA said in a statement.

“This in turn successively triggered a premature release of the parachute and the backshell [heat shield], a brief firing of the braking thrusters and finally activation of the on-ground systems as if Schiaparelli had already landed. In reality, the vehicle was still at an altitude of around 3.7km (2.3 miles).”

The €230m ($251m) Schiaparelli had spent seven years travelling 496m kilometres (308m miles) onboard the so-called Trace Gas Orbiter to within a million kilometres of Mars when it set off on its own mission to reach the surface.

Source: https://www.theguardian.com/science/2016/nov/24/mars-lander-smashed-into-ground-at-540kmh-after-misjudging-its-altitude

NASA’s Mars Reconnaissance Orbiter High Resolution Imaging Science Experiment (HiRISE) imaged the ExoMars Schiaparelli module’s landing site on 25 October 2016, following the module’s arrival at Mars on 19 October.

The zoomed insets provide close-up views of what are thought to be several different hardware components associated with the module’s descent to the martian surface. These are interpreted as the front heatshield, the parachute and the rear heatshield to which the parachute is still attached, and the impact site of the module itself.

In the image, north is up; west to the left. Schiaparelli was travelling from west to east. The image scale is 29.5 cm/pixel. The brightness of the individual zooms have been adjusted to best reveal the features against the martian surface in each case.

The 100 m scale bar in the main image is only indicative, as the HiRISE image was taken at an oblique angle. The distances given between the various components in the main text have been corrected for this effect.

Image source: http://www.esa.int/spaceinimages/Images/2016/10/Zooming_in_on_Schiaparelli_components_on_Mars

http://sicv.activearchives.org/toolbox/Zooming_in_on_Schiaparelli_components_on_Mars.html

Texture mapping

3a_img_4460web_toolbox02

Code: http://gitlab.constantvzw.org/SICV/toolbox