They have the same chocolate coloured hair, they share dark, almond-shaped eyes and have a similar impish smile.
Since birth their family, friends and teachers have mixed them up. When their high school prom dates came to their house in a small town in east Texas, the two boys could not tell which twin they were there to collect.
Yet despite their similarities, certain features do mark them apart. “We feel that we are more different than others perceive us to be,” explains Erin. “There are subtle differences in our faces – Audri has a slightly rounder face, maybe because she smiles more. Our lips also have a slightly different contour, but it is only noticeable if we are side by side in a picture.”
Those subtle cues are often missed by facial recognition software, which uses computers to automatically detect distinctive features and match them to images held in vast databases. And it is for this reason that Erin and Audri are now helping computer scientists, such as Kevin Bowyer at the University of Notre Dame in Indiana, to put their facial recognition and other “biometric identification” techniques to the ultimate test.
Along with his colleague Patrick Flynn, also at the University of Notre Dame, Bowyer recently went to the annual Twins Day Festival in Twinsburg Ohio to see how well some of the latest technologies can do. “The fraction of identical twins in the population is slowly growing over time so it is potentially a real issue that you would want to distinguish between twins,” says Bowyer.
“Fingerprints are a pretty good way of telling identical twins apart, but not quite as good as it is in the general population.”
“Facial recognition is not so good. It is something our lab is looking at – whether moles, facial scars and other marks can be used to distinguish people who look very much alike. At the moment iris recognition is the best method we have for easily distinguishing twins.”
This sort of work, along with attempts to find new ways of identifying people, is revealing that we may differ from our neighbours in some rather surprising ways. Here are some of the most unusual but utterly unique features of your body.
For most of us, our ears are merely gristly appendages sticking out from the sides of our heads. They are often too small, too big, too prominent or just too weird looking for our liking.
But perhaps it is time to give our ears a little more respect. The complex patterns of ridges and furrows formed by the cartilage of your outer ears are remarkably unique to you.
It is possible to recognise someone from even a fuzzy picture of their ear with 99.6% accuracy
Since the 1950s, forensic scientists been using measurements of suspects’ ears to match them to prints left behind at the scenes of crimes. Then in 1998, researchers showed that computers could accurately distinguish between images of ears.
This has led to 3D scanning techniques that build up a detailed model of the ear and new methods that extract distinctive geometric features of the ear. By analysing a combination of features, such as the distance from the lobe to two points on the upper edge, it is possible to recognise someone from even a fuzzy picture of their ear with 99.6% accuracy in just 0.02 milliseconds.
“As you grow older your ear doesn’t change – it just grows larger but the proportions remain the same,” explains Mark Nixon, a computer vision expert at the University of Southampton who has been one of the leading researchers in ear recognition.
But the individuality of our ears does not stop there – they sound different too. The microscopic hair cells in the cochlea of our inner ears, which detect soundwaves and turn them into minute electrical signals for our brain, also produce their own faint sounds.
This occurs as a byproduct of the process that amplifies quiet sounds, which causes these cells to vibrate and produce a noise that is inaudible to the human ear, but can be detected using sensitive microphones. These are known as octoacoustic emissions and are already used by doctors to test the hearing of young children by playing a series of clicks to see how the hair cells respond. But researchers have found that in each of us they are subtly different.
Each person’s hair cells produce a slightly different noise, which is further changed as the sound travels through the bones in the inner ear, the eardrum and along the ear canal.
It raises the possibility of being able to unlock your phone simply by lifting it to your ear or putting in a pair of headphones. Indeed, technology manufacturer NEC announced last year that it had developed an earphone with a built in microphone that can extract the sounds that resonate within ear cavities to identify individuals with greater than 99% accuracy.
YOUR BODY ODOUR
Like it or not, you smell. Underlying the collection of perfumes, deodorants and other scents we slather on ourselves to seem more attractive is a distinct natural body odour.
Studies funded by the US Army Research office have found it is possible to identify individuals by looking at the fog of chemicals we leave in our wake as we move around. Scientists at the University of Bristol showed they could distinguish individuals by looking at a combination of 44 compounds secreted by the body, which evaporate off to form a ‘thermal plume’ that surrounds us.
The UK’s Forensic Science Service has also examined the potential of using the scent in traces of sweat left at crime scenes
It is this unique scent trail that search dogs are able to detect and follow, and researchers have been attempting to use “electronic noses”, which sniff out the chemical constituents of an odour, to do something similar. The UK’s Forensic Science Service has also examined the potential of using the scent in traces of sweat left at crime scenes during investigations.
Spanish company Ilía Sistemas has developed a technique that is able to identify members of a group from their body odour with an accuracy of 85%.
Most recently, researchers at the University of Information Science and Technology St Paul the Apostle in the Republic of Macedoniaand Kyushu University in Japan, claim they can distinguish between individuals with 100% accuracy, using equipment that produces atomic “fingerprints” of a scent and then running it through an artificial neural network on a computer.
Sunil Jha, a bioinformatics engineer who led that work, says odour could eventually be used as a form of identification in biometric passports. “Body odour has great potential to be used in biometric applications,” he claims. “I’m sure our approach can discriminate between the pure body odour and the odour from deodorants for example.”
THE WAY YOU WALK
Never has Phil Collins sung truer words: “The only thing about me is the way I walk.” The idea of identifying people from the way they move has been around for decades and has become a popular idea in the security services.
While putting one foot in front of another might seem like a simple act, each of us has tiny ticks, bounces and ways of swinging our legs that are unique.
It ultimately stems from subtle differences in our bone structure, muscles and our sense of balance. For most people, these subtleties are difficult to spot with the human eye, but for machines they are telltale signs of individuality. Using computers to break down movement into a series of points or more recently, build 3D kinetic models of the way someone as they walk, it is possible to take a video of someone moving and match it to a database with astonishing accuracy.
Mark Nixon at Southampton University was one of the first researchers to do this and found that while the way we walk is often quite individual, the way we run is even more so. “Walking is a very passive motion while running is a forced motion and so it much more distinctive,” he explains. “Computers can break down the way you move into a set of numbers that are unique to you. We can identify you from the way your walk, the way you run and the way you go from walking to running.”
Mobile phone manufacturers are now developing technology that can detect users from the way they move around. One presented at the Biometrics 2016 conference in London last month requires the user to write a code in the air with their phone in their hand and its onboard sensors detects specific eccentricities of their movement to confirm their identify. Others are hoping that mobile phones will be able to detect they are with their registered user simply by monitoring the way they move around while it is in their pocket.
But there are also problems. “What would happen if you had changed your shoes or wore a tighter pair of pants,” warns Bowyer. “If it is raining you might also walk differently on grass than you would on concrete.”
Some of us have pert posteriors while others could probably do with a little toning, but your bottom is actually quite remarkable.
According to researchers in Japan, they can tell who you are purely from your backside.
They have developed a pressure-sensitive pad that can be fitted into the seat of a car or a desk chair. A total of 360 sensors detect the contours of a person’s backside when they sit down, differences in where they apply pressure on the seat – they might sit more on one cheek than the other for example – and how much of their body touches the seat. So slouchers will have a different sitting profile from those that sit up straight.
This information from the sensors is then used to match against a profile held in a computer database. Shigeomi Koshimizu, from the Advanced Institute of Industrial Technology in Tokyo, Japan, believes his system could be used an anti-theft device in cars or to allow workers to unlock their computers simply by sitting down rather than having to remember complex passwords.
Put your fingers in your ears, and keep them there while placing your elbows on your desk or the arm of the chair you are sitting in.
The chances are you will probably hear a low humming noise coming up through your arms from the room around you that disappears when you lift your elbows again. What you have just experienced is your skull conducting the sound vibrations from your surroundings directly to your inner ear.
Bone is remarkably good at conducting sound in this way. But the size of your skull, the amounts of soft tissue around it and the hollow cavities it contains, means the way it vibrates is different in every person. Researchers have found that by playing a short clip of white noise and then listening to the way it is changed by a person’s skull, it is possible to pinpoint their identity with 97% accuracy.
We are carrying around our own personal tuning forks inside our heads
Essentially it means we are carrying around our own personal tuning forks inside our heads.
The researchers’ SkullConduct system could be used to unlock mobile phones or other devices like Google Glass in the future by simply playing a sound into your skull and listening to what happens.
“We send a known signal into the skull and analyse what ‘comes out’,” explains Andreas Bulling, an engineer at the Max Planck Institute for Informatics in Saarbrücken, Germany, who led the research. “It could potentially also be used for other parts of the skeleton too.”
While the loops and whirls of skin on the pads of your fingers are perhaps one of the best known forms of identification, it seems another set of unique features have been staring us in the face every time we use a fingerprint to unlock our smartphone.
Hidden within the shiny surface of our fingernails are a series of ridges and contours that have been found to be unique in studies of thousands of samples. When examined in the right light, the patterns on the surface of the nail plate – the hard part of the nail, which is made of translucent keratin protein – can be revealed.
Researchers at Indian Institute of Technology in New Delhi even showed they were able to see these distinct patterns in photographs of fingernails when people were wearing nail polish. They claim by taking scans of the nails from three fingers – the ring, middle and index – they can reliably identify individuals.
But fingernails also grow and replace over time, meaning these patterns can change every six months or so. Instead, some researchers have decided to focus on the layers of skin that lie beneath the nail.
The nail bed, as this delicate pinkish tissue is known, has distinct channels that an outer layer of cells slot into like a tongue and groove joint used in woodwork. The spacing of these grooves remain proportionately the same throughout your life but have different patterns in each person.
One company developed a system called Nail-ID that used low powered lasers to measure these in an attempt to develop a new way of confirming a person’s identity.
THE PORES ON YOUR NOSE
Splattered across your nose like seeds covering a strawberry are hundreds of tiny pores. These are the enemy of beauticians everywhere due to their tendency to gather grime and clog.
But regardless of whether you have the flawless skin of a supermodel or a face that oozes grease like a deep-fat frier, your nose pores help to distinguish you from others. Researchers at Shangdong University in China, along with colleagues at Kinki University in Japan, insist the distribution of pores on your nose is stable throughout your life.
They have developed a system that can recognise nose pores from photographs and match these to a database. They claim they could identify people taking part in their study with 88% accuracy based purely on the pores of their nose.
While esoterically fascinating, there could also be real benefits to our uniqueness. Most of us struggle to remember the countless passwords and pin numbers we need to access our computers, phones and services online.
Repeated hacks of major companies have also shown just how vulnerable those passwords can also be. It is clear we need other ways to authenticate our identity. Fingerprint sensors have now become common on new mobile phones but already some hackers have found ways to get around these.
Instead experts believe we need a mixture of several different biometric features to help us prove our identity. There are dozens of options – the way we type, our heartbeats and even the proteins in our hair are all unique to us as individuals. “We have always argued that biometrics will make your life more convenient and secure than using passwords,” says Mark Nixon.
“This is because your body is unique. What we are finding is that it is even more unique than we realised.”
Erin and Audri Nelson are already finding other ways of telling themselves apart that could help those around them. “We used to dislike the stares when we were little but now we embrace it,” says Audri. “We will even try harder to look alike with matching outfits, similar haircuts and makeup. But if you look closely, the way we walk is very different.”