Pretty trippy test. I only got 1 out of 3 (the 3rd one). Maybe someone can post the actual test videos but I don't know how.
http://news.bbc.co.uk/2/hi/science_a...t/10284925.stm How blind to change are you?
By Victoria Gill
Science reporter, BBC News
Have you ever failed to notice a friend's radical new haircut? Or missed a road sign showing a change in the speed limit?
This failure to notice what should be very apparent is something we unconsciously experience every day as our brains filter the barrage of visual information which we are flooded with. And apparently it has a name; it is called change blindness.
Scientists at Queen Mary, University of London, have invented a unique spot-the-difference-style computer game in order to study it.
Milan Verma, a scientist at Queen Mary, explains: "It's the phenomenon where seemingly striking or obvious changes are not noticed." He and his colleagues are asking volunteers to play the game - which involves looking at a screen as it flashes between two images of the same scene.
"It flicks between a pre-change version and a post-change version of the scene," Dr Verma explains. "The volunteers simply have to press the button and tell us exactly when they spot the change."
Trying out the game at Dr Verma's office, my initial reaction was self-satisfaction; I spotted the difference in the first scene - a picture of a butterfly with orange stripes on its wings - almost immediately.
In the pre-change scene the colourful insect had two stripes - one on each wing, and on the post-change, there was just one. Easy. Next?
But I was quickly reminded that I am just as "change blind" as the next person. As an image of an iceberg scene with five penguins on it flashed in front of me, I stared blankly, unable to see a difference.
"I'll let you off - there is a lot going on in this image," Dr Verma reassured me. "But it's quite a big change."
He had to give me a clue - directing me to the area of the image where the change occurred - before I realised that a whole chunk of iceberg was missing in the post-change image.
That represented one of the fundamental factors about change blindness; a whole chunk of iceberg might seem like an easier thing to spot than the stripe on a butterfly wing, but it is not as obvious to the human brain.
"The butterfly image is easy because the changed scene violates our expectations," explains Dr Verma. "We expect butterflies to be symmetrical - to have two identically marked wings - so one that isn't really stands out to us."
Neuroscientists, as well as developers of artificial intelligence, have been interested in this facet of human perception for many years. In fact, the Queen Mary team incorporate their biological findings into the design of robots - studying the basis of human vision and perception in order to artificially recreate it.
And Dr Verma says this might be the first truly unbiased scientific study of change blindness.
"Previously, scientists have studied this by manually manipulating pictures," he said.
"So they'd use... image manipulation software, make a deliberate change and then ask viewers: 'Can you see the change, yes or no?'."
This, he says, is cheating. If a human scientist makes a change to a picture, they are making a very human decision about what and where that change is - choosing to remove the bird from the corner of the park view, or to change the colour of the sofa in a living room scene.
"So they're making some subjective judgement about how noticeable they think the change is."
Artificial intelligence
In this study, Dr Verma and his colleague and supervisor, Professor Peter McOwan, created an algorithm that meant the computer "decided" how to change the image.
Professor McOwan told BBC News: "This is, as far as I'm aware, the first time ever that artificial intelligence [AI] technology has been used to generate experimental stimuli to test human perception.
"It brings together two interesting fields of study- AI and human visual intelligence."
Dr Verma and Professor McOwen designed software that underlies the game's ability to make a change to each image.
Dr Verma describes this as a "genetic algorithm". It essentially tells the computer to change the images in a process akin to evolution.
"It's like a process of survival of the fittest," explained Dr Verma. "Darwin suggested that a fit individual is one that can best survive in its surroundings - like a moth that can camouflage with the bark of a tree."
But in this case "fitness" is determined by the smallest difference between the pre- and post-change scenes, in terms of how attention-grabbing they are.
The computer uses information about human attention and perception to generate two pictures that a person will view in exactly the same way; two images that are equally attention-grabbing. This means the scientists get an accurate measure of how noticeable the change is and there is no "human bias" of the results.
The research is beginning to reveal where in a scene people direct their attention as well as what kinds of changes are more noticeable.
"It all boils down to contrast," he says. "So colour or orientation contrast; luminance contrast in terms of light and dark things that pop out.
"And it's what's easy to spot in terms of our viewing attention behaviour.
"So when we walk into a room, our eyes are attracted to a particular region and because we're attending to that region, if there was a change made there, we would perhaps notice it more quickly."
The scientists have published their findings and described their unique approach in the journal Vision.
Their test has also revealed some more specific points about what types of contrast are more obvious to us. We are more likely, for example, to spot that an object has been removed from a scene than if it has changed colour.
Attention-grabbing
The team has already had interest from companies that want to apply the findings to the design of safety notices and advertising displays - to grab our attention.
And there is potential for these results to be used in more clandestine ways. Just like a magician might use misdirection, police or security services might take advantage of things that make people look.
"You can use these attention-grabbing principles to, for example, direct someone's attention to a particular spot and that could be a spot where there is a camera," explains Dr Verma. "So you could take a photograph - a frontal image - that could be useful for police or security services."
Billboards in Times Square, New York City Brands and adverts aim to stand out and grab our attention
As well as applying these findings to the development of intelligent robots, Professor McOwan is more informally interested in the role of change blindness in magic.
"There has some interest recently at looking at how magicians misdirect people [with these attention-grabbing methods]," he says. "They have a role in the 'now you see it now you don't' tricks."
There are also more lucrative roles for these findings - in advertising or website design.
And Dr Verma points out that they have already been applied by the emergency services.
"You might notice that they are starting to use different kinds of beacons and lights and different, very strange-sounding sirens," he says.
"This is because the classic siren sound has been so over-used by the media that it just doesn't stand out to us in the same way any more."
So apparently, we all have an excuse for not noticing a friend's new look. It's a fundamental part of our perception - we are simply too focused on looking them in the eye and on what they have to say to notice something so superficial.
At least, from now on, that will be my excuse.
You are currently viewing our boards as a guest which gives you limited access to view most discussions and access our other features. By joining our free community you will have access to post topics, communicate privately with other members (PM), respond to polls, upload content and access many other special features. Registration is fast, simple and absolutely free so please, join our community today!
The banners on the left side and below do not show for registered users!
