NeuralHash is the perceptual hashing model that back’s Apple’s new CSAM (child sexual abuse material) reporting mechanism. It’s an algorithm that takes an image as input and returns a 96-bit unique identifier (a hash) that should match for two images that are “the same” (besides some minor perturbations like JPEG artifacts, resizing, or cropping).
Yesterday, news broke that researchers had extracted the neural network Apple uses to hash images from the latest operating system and made it available for testing the system. Quickly, artificially colliding adversarial images were created that had matching NeuralHashes. Apple clarified that they have an independent server-side network that verifies all matches with an independent network before flagging images for human review (and then escalation to law enforcement).
Posts tagged perception
One of the richest veins in temporal-perception research is on the effect of emotion on cognition, and Droit-Volet has conducted a number of compelling studies that explore the relationship. In a recent series of experiments, her subjects viewed a series of images of faces, each of which was neutral or expressed a basic emotion, such as happiness or anger. Each image lasted onscreen for anywhere from 0.4 seconds to 1.6 seconds, and the viewer was asked to say whether the image lasted for a “short” or a “long” time—that is, closer to one of the two standard durations they’d been trained beforehand to recognize. Consistently, viewers reported that happy faces seemed to last longer than neutral ones, and both angry and fearful faces seemed to last longer still.
She starts with a clip that’s been digitally altered to sound like jibberish. On first listen, to my ears, it was entirely meaningless. Next, Das plays the original, unaltered clip: a woman’s voice saying, “The Constitution Center is at the next stop.” Then we hear the jibberish clip again, and woven inside what had sounded like nonsense, we hear “The Constitution Center is at the next stop.” The point is: When our brains know what to expect to hear, they do, even if, in reality, it is impossible. Not one person could decipher that clip without knowing what they were hearing, but with the prompt, it’s impossible not to hear the message in the jibberish. This is a wonderful audio illusion.
Most displays are looking to play things faster. We’ve got movies at 60 frames per second, and gaming displays that run at 144 fps. But what about moving in the other direction? [Bryan Boyer] wanted to try this out, so he built the VSMP, or Very Slow Movie Player. It’s a neat device that plays back a movie at about 24 fph (frames per hour) on an e-ink display to demonstrate something that [Bryan] calls Slow Seeing, which, he says “helps you see yourself against the smear of time.” A traditional epic-length movie is now going to run you greater than 8,000 hours of viewing.
“Most displays are looking to play things faster. We’ve got movies at 60 frames per second, and gaming displays that run at 144 fps. But what about moving in the other direction? [Bryan Boyer] wanted to try this out, so he built the VSMP, or Very Slow Movie Player. It’s a neat device that plays back a movie at about 24 fph (frames per hour) on an e-ink display to demonstrate something that [Bryan] calls Slow Seeing, which, he says “helps you see yourself against the smear of time.” A traditional epic-length movie is now going to run you greater than 8,000 hours of viewing.“
Our notions of what it means to have a mind have too often been governed by assumptions about what it means to be human. But there is no necessary logical connection between the two. There is often an assumption that a digital mind will either be, or aspire to be, like our own. We can see this at play in artificial beings from Pinocchio to the creature in Mary Shelley’s Frankenstein to 2001: A Space Odyssey’s HAL to Data from Star Trek: The Next Generation. But a machine mind won’t be a human-like mind — at least not precisely, and not intentionally. Machines are developing a separate kind of interaction and interrelation with the world, which means they will develop new and different kinds of minds, minds to which human beings cannot have direct access. A human being will never know exactly what it’s like to be a bot, because we do not inhabit their modes of interaction.
navigating the limited piece of physical reality we encounter in life, and remaining mentally and emotionally secure enough to survive, find mates, and propagate the species, requires an unquestioning, and when you think about it, strikingly unreasonable confidence in ourselves and in the world. Since full awareness of reality as-it-is was not an option for our ancient ancestors (as the overwhelm caused by so much data would have diminished, rather than enhanced, their chances of survival), evolution equipped them –and, as their descendants, us too — with brains capable of generating a convincing illusion of the reality of our own small words.
“our interpretations of the world are necessary tools for making sense of it, but fixing onto these beliefs as if they revealed the true nature of things can be dangerous. The world is always more chaotic than the order our closed systems of signs impose.”
–Brent Ables (on the work of Umberto Eco)
“Yet tangles remain tangles, even for those inextricably tangled within them, and the greatest tangles of all are still only very partially seen.”
–Nick Land, Templexity
As interpreted, ‘Real’ Reality is something that sits outside of ‘Official Reality’. Official or ‘Red Reality’ is the reality of mainstream culture which is the preferred reality of ‘Power’ (substitute Power for Ruling Archon as is your prerogative). It is through the construction of this Official Reality that allows ‘Power’ to govern. Within the Red sphere of Reality ‘Power’ can be said to play by its own rules. The diagram also suggests that there is an expanded ‘Reality’ within which you can play by different rules. It is at the the boundary between the official sphere of reality and the outside that ‘Power’ gets to choose which rules and which cards are in and out of play.
Is “now” expandable? Why do you seem to experience time in slow motion in a sudden emergency, like an accident? Eagleman’s (terrifying) experiments show that in fact you don’t perceive more densely, the amygdala cuts in and records the experience more densely, so when the brain looks back at that dense record, it thinks that time must have subjectively slowed down, but it didn’t. “Time and memory are inseparable.” This also explains why time seems to speed up as you age. A child experiences endless novelty, and each summer feels like it lasted forever. But you learn to automatize everything as you age, and novelty is reduced accordingly, apparently speeding time up. All you have to do to feel like you‘re living longer, with a life as rich as a child’s, is to never stop introducing novelty in your life.
“Today’s men’s nerves surround us. Each technological extension gone outside is electrical involves an act of collective environment. The human nervous environment system itself can be reprogrammed with all its private and social values because it is content. He programs logically as readily as any radio net is swallowed by the new environment. The sensory order.”
–William S. Burroughs, picking up a copy of The Nation.
The ability to remember the past and imagine the future can significantly affect a person’s decisions in life. Scientists refer to the brain’s ability to think about the past, present, and future as “chronesthesia,” or mental time travel, although little is known about which parts of the brain are responsible for these conscious experiences. In a new study, researchers have used functional magnetic resonance imaging (fMRI) to investigate the neural correlates of mental time travel and better understand the nature of the mental time in which the metaphorical “travel” occurs.
One of the challenges of neural networks is understanding what exactly goes on at each layer. We know that after training, each layer progressively extracts higher and higher-level features of the image, until the final layer essentially makes a decision on what the image shows. For example, the first layer maybe looks for edges or corners. Intermediate layers interpret the basic features to look for overall shapes or components, like a door or a leaf. The final few layers assemble those into complete interpretations–these neurons activate in response to very complex things such as entire buildings or trees. One way to visualize what goes on is to turn the network upside down and ask it to enhance an input image in such a way as to elicit a particular interpretation. Say you want to know what sort of image would result in %E2%80%9CBanana.%E2%80%9D Start with an image full of random noise, then gradually tweak the image towards what the neural net considers a banana
The results show that not all types of spookiness emerge in the same way from the brain. “They show that the neural networks involved in the feeling of a presence are not the same as those involved in out-of-body experiences or in seeing a doppelgänger,” says the lead author of the study, cognitive neuroscientist Olaf Blanke of the Swiss Federal Institute of Technology of Lausanne (EPFL).
Relaxing the null hypothesis makes for great storytelling, but it’s an unsettling way to live. Which information and which perspectives we take into account, when we try to decide whether a pattern we’ve matched is real or an apophenic false alarm, affects our ability to determine whether something has gone away or whether we’ve just stopped believing in it. In the praxis of science we try to keep the false alarm rate down with things like study size and statistical inference and meta-analyses and replication, and we still get it wrong a lot.
Umpire Decision Review System (DRS), a suite of technologies which assist - or rather, overrule - the umpire adjudicating some of sports greatest unknowables, the LBW, and the snick. Of course, these technologies, intended to increase accuracy, only inflamed controversy as their own accuracy was questioned as much as the human umpires. LBW is, after all, an epistemological problem - the question of whether a ball which strikes the batsman would have struck the wicket were the batsman not there is a question for Plato, not for machines.
Enter the EyeWire project, an online game that recruits volunteers to map out those cellular contours within a mouse’s retina. The game was created and launched in December 2012 by a team led by H. Sebastian Seung, a neuroscientist at the Massachusetts Institute of Technology in Cambridge. Players navigate their way through the retina one 4.5-micrometer tissue block at a time, coloring the branches of neurons along the way. Most of the effort gets done in massive online competitions between players vying to map out the most volume. (Watch a video of a player walking through a tissue block here.) By last week, the 120,000 EyeWire players had completed 2.3 million blocks. That may sound like a lot, but it is less than 2% of the retina.
Photography has become so fundamental to the way we see that “photography” and “seeing” are becoming more and more synonymous. The ubiquity of photography is, perhaps ironically, a challenge to curators, practitioners, and critics. Why look at any particular image, when they are literally everywhere? Perhaps “photography” has become so all-pervasive that it no longer makes sense to think about it as a discreet practice or field of inquiry. In other words, perhaps “photography,” as a meaningful cultural trope, is over.
Photography does not lend itself to defamiliarization easily, thus making it the unlikeliest of all art forms. As it happens, the challenge plays out on both sides of the process, for photographers and viewers. What happened to be in front of a camera lens can be found depicted in the resulting photograph. However, given the process itself and its myriad of choices, the photograph is little more than a manipulated two-dimensional representation of what previously existed in four dimensions (three spatial, one – often forgotten – time).
We know that applications such as Google Maps, Google Earth and StreetView already acquiesce to regulations that require obscuration of government installations, private companies’ facilities in some cases, some brands, and private citizens faces and number plates—even as it works hard to decipher items like house numbers. In other words, technology is used to differentiate what we can see and not see, depending on the legal or ethical (or otherwise) standards of a particular place. For the most part, Google Maps, Google Earth and StreetView are forms of augmented reality. They digitally render reality with forms of markup, of contextual data, which adds to our perception of places. Except in the cases of blur, pixelation and, it could be argued, accidental presentation of various kinds of render ghosts—people and things only partly captured or partly presented, artifacts of digital accident or persistent memory. Some kind of determination is made that there are things present in reality that we can’t or shouldn’t see.
The human brain, we have learned, adjusts and recalibrates temporal perception. Our ability to encode and decode sequential information, to integrate and segregate simultaneous signals, is fundamental to human survival. It allows us to find our place in, and navigate, our physical world. But music also demonstrates that time perception is inherently subjective—and an integral part of our lives. “For the time element in music is single,” wrote Thomas Mann in his novel, The Magic Mountain. “Into a section of mortal time music pours itself, thereby inexpressibly enhancing and ennobling what it fills.”
In the same way that the transition from film to digital is now taken for granted, the shift from cameras to networked devices with lenses should be obvious. While we’ve long obsessed over the size of the film and image sensors, today we mainly view photos on networked screens—often tiny ones, regardless of how the image was captured—and networked photography provides access to forms of data that go beyond pixels. This information, like location, weather, or even radiation levels, can transform an otherwise innocuous photo of an empty field near Fukushima into an entirely different object. If you begin considering emerging self-metrics that measure, for example, your routes through cities, fitness level, social status, and state of mind (think Foursquare, Nike+, Facebook, and Twitter), you realize that there is a compelling universe of information waiting to be pinned to the back of each image. Once you start thinking of a photograph in those holistic terms, the data quality of stand-alone cameras, no matter how vast their bounty of pixels, seems strangely impoverished. They no longer capture the whole picture.
Over the centuries, the illusion of mastering time through obedience to it came into acceptance. Across Europe, the medieval monastery’s bell tolled as a reminder to eat, sleep and pray. But while there must have been some soul’s release in relinquishing earthly sovereignty to that sound, as the clock’s authority spread, we sealed all the gaps through which curiosity might seep into our days. Curiosity, after all, could lure the susceptible way off track, as the Italian poet Petrarch learned in the spring of 1336, when he famously climbed Mont Ventoux, motivated by “nothing but the desire to see its conspicuous height.” One of the texts he carried along was Saint Augustine’s “Confessions,” detailing the moral dangers of such expeditions, when men “go out to admire the mountains,” or the course of the stars, and therein forget themselves. Chastened, Petrarch made his descent in silence.
clicks-ReflectanceCP018VRgb0.55-crop (via http://www.pawfal.org/dave/blog/2013/11/project-nightjar-camouflage-data-visualisation-and-possible-internet-robot-predators/)
A warning for future space colonizers: Babies born in space might not ever figure out how to deal with gravity. Jellyfish babies, at least, have to deal with massive vertigo on Earth after spending their first few days in space. NASA first started sending jellyfish to space aboard the Columbia space shuttle during the early ‘90s to test how space flight would affect their development. As cool as being an astronaut baby sounds, the jellies didn’t develop the same gravity-sensing capabilities as their Earthly relatives.
Some of this fear results from imperfect risk perception. We’re bad at accurately assessing risk; we tend to exaggerate spectacular, strange, and rare events, and downplay ordinary, familiar, and common ones. This leads us to believe that violence against police, school shootings, and terrorist attacks are more common and more deadly than they actually are – and that the costs, dangers, and risks of a militarized police, a school system without flexibility, and a surveillance state without privacy are less than they really are.
The central, unifying theme of the institute was time. Not physical time, but biological and psychological time. How does our brain perceive physical time? What is the structure of perceived time? What regulates biological oscillations in humans, animals and even algae? Can environmental cues modify temporal perception? The close proximity of so many disciplines made for fascinating coffee-break discussions, forcing us to re-evaluate our own research findings in the light of the discoveries made in neighboring labs and inspired us to become more creative in our experimental design.
Curiosity does not seem to be a fundamental drive, unlike what I am told are the three basic biological drives (seeking pleasure, avoiding pain and conserving energy), so it is probably derived. Curiosity requires a certain energy surplus, since its visible signature is a restless dissipation of energy, but it does not seem directly motivated by energy conservation concerns. So is it derived from pleasure-seeking or pain-avoidance or some mix of the two? Does that make a difference?
A lot of psychological research has tried to make sense out of security, fear, risk, and safety. But however fascinating the academic literature is, it often misses the broader social dynamics. New York University’s Harvey Molotch helpfully brings a sociologist’s perspective to the subject in his new book Against Security.
Robbins, who is thirty-eight and lives in Las Vegas, is a peculiar variety-arts hybrid, known in the trade as a theatrical pickpocket. Among his peers, he is widely considered the best in the world at what he does, which is taking things from people’s jackets, pants, purses, wrists, fingers, and necks, then returning them in amusing and mind-boggling ways. Robbins works smoothly and invisibly, with a diffident charm that belies his talent for larceny.
Color input is processed at every stage of the nervous system, starting from the retina, and the experience of color is affected by numerous contextual and historical factors. Nevertheless, in broad strokes we can describe a color by three numbers. A common trope in science and math is that if we can describe an object by a set of numbers, we can think of these numbers as its coordinates in an abstract space, with each point in the space corresponding to a different object, and we can study the geometry of the space
So what I’m going to talk today, obliquely, about is a project that I’ve been sort of accidentally engaged in for the last six months or so, to which I gave the name “The New Aesthetic,” which is a rubbish name but it seems to have taken hold. And people are responding to it, which is good. And I’m going to try and talk through some of the symptoms of that, this project, this way of seeing, that is itself about ways of seeing. And this talk is about the aesthetics of that. So this idea extends in all directions and through all forms in media and technologies. But because I have nice big screens here, I’m going to show you a lot of pictures of it.