Researchers believe they can diagnose depression by looking at Instagram photos

It may seem that everyone on Instagram is living their best life, but according to new research photos that are bluer, darker, and less populated might reveal a struggle with mental health.

Researchers from the University of Vermont and Harvard University, have found that they could diagnose depression by using machine learning to study the composition of Instagram photos.

According to the researchers, photos that are bluer, more gray, dark or with fewer faces indicate that the profile’s owner has depressive tendencies.

The computer’s detection rate of 70% is more reliable than the 42% success rate of general-practice doctors diagnosing depression in-person.

“This points toward a new method for early screening of depression and other emerging mental illnesses,” says Chris Danforth, a professor at the University of Vermont who co-led the new study. “This algorithm can sometimes detect depression before a clinical diagnosis is made.”

“So much is encoded in our digital footprint. Clever artificial intelligence will be able to find signals, especially for something like mental illness.”

Featured image courtesy of Ink Drop / Shutterstock.com.

To conduct the study, the researchers used the Instagram feeds of 166 people, around half of whom had reported having been clinically depressed in the last three years.

The researchers then collected and analysed 43,950 photos, using insights from well-established psychology research, about people’s preferences for brightness, colour, and shading. They also investigated the filters used by healthy and mentally ill people.

They found that healthy individual chose Instagram filters, like Valencia, that gave their photos a warmer brighter tone. Among depressed people the most popular filter was Inkwell, making the photo black-and-white.

Faces in photos also turned out to provide signals about depression. The researchers found that depressed people were more likely than healthy people to post a photo with people’s faces—but these photos had fewer faces on average than the healthy people’s Instagram feeds.

“People suffering from depression were more likely to favour a filter that literally drained all the colour out the images they wanted to share,” said Danforth and Andrew Reece of Harvard University who co-wrote the study.

“Fewer faces may be an oblique indicator that depressed users interact in smaller settings.”

Images courtesy of Reece and Danforth

More than half of a general practitioners’ depression diagnoses are false, but the computational algorithm was able to achieve a detection rate of 70%, while GPs are said to have only a 42% success rate.

The new study also shows that the computer model was able to detect signs of depression before a person’s date of diagnosis. “This could help you get to a doctor sooner,” Danforth says. “Or, imagine that you can go to doctor and push a button to let an algorithm read your social media history as part of the exam.”

Crypto-currency mining is hindering the search for alien life

Researchers searching for extraterrestrial life are struggling to get the computer hardware they need, due to crypto-currency mining. "We'd like to use the latest GPUs...and we can't get 'em," said Dan Werthimer. Demand for GPUs has soared recently thanks to crypto-currency mining.

Source: BBC

Genetic study of soil reveals new family of antibiotics

Researchers have discovered a new family of antibiotics in samples of soil. In their paper, published in the journal Nature Microbiology, the group describe how the antibiotics, named malacidins, have been shown to kill superbugs such as MRSA, which are resistant to current antibiotics.

Source: Phys.org

UK police are identifying suspects using fingerprint scanners

UK police have begun using a mobile fingerprinting system to identify people in less than a minute. Fingerprints are compared against 12 million records stored in the national criminal and immigration fingerprint databases and, if a match is found, return info like the individual’s name and date of birth.

Source: Wired

Robots 1,000 times smaller than a human hair could treat cancer

Scientists from Arizona State University and The Chinese Academy of Sciences just figured out how to build tiny robots that travel through the body's blood stream, hunting for tumors, without doing any harm to healthy cells along the way. In tests on mice, average survival times doubled.

NASA is bringing back Cold War-era rockets to get to Mars

NASA is planning to use atomic rockets to help get humans to Mars. Unlike conventional rockets that burn fuel to create thrust, the atomic system uses the reactor to heat a propellant like liquid hydrogen, which then expands through a nozzle to power the craft, doubling the efficiency at which the rocket uses fuel.

Source: Bloomberg

Facial recognition systems have gender and racial biases

Research conducted by Massachusetts Institute of Technology and Stamford University has found that AI-powered facial recognition systems have gender and racial biases, which means because they have been trained using large data sets of white males they are better at picking out that group than any other.

Source: The Inquirer

Skydio unveils its obstacle-dodging, thrill-seeking, AI-powered drone

An autonomous drone startup founded by former MIT researchers has today launched its R1, a fully autonomous flying camera that follows its subjects through dense and challenging environments.

In a promotional video, launched to introduce the autonomous camera, R1 can be seen following an athlete as she parkours her way through dense woodland.

The drone’s makers Skydio have explained that the camera combines artificial intelligence, computer vision, and advanced robotics and works by anticipating how people move, so R1 can make intelligent decisions about how to get the smoothest, most cinematic footage in real-time.

“The promise of the self-flying camera has captured people’s imaginations, but today’s drones still need to be flown manually for them to be useful,” said Adam Bry, CEO and co-founder of Skydio.

“We’ve spent the last four years solving the hard problems in robotics and AI necessary to make fully autonomous flight possible. We’re incredibly excited about the creative possibilities with R1, and we also believe that this technology will enable many of the most valuable drone applications for consumers and businesses over the coming years.”

Launching today is the Frontier Edition of R1, which is aimed at athletes, adventurers, and creators.

This version of R1 is powered by the Skydio Autonomy Engine, enabling it to see and understand the world around it so that it can fly safely at speeds of upto 25mph while avoiding obstacles.

The autonomous drone is fitted with 13 cameras, which gives it the ability to map and understand the world in real-time, allowing it to be fully autonomous and independently capture footage that in Skydio’s words “once required a Hollywood film crew” and will “enable a new type of visual storytelling”.

The R1 “Frontier Edition” is available for order now on Skydio’s website for $2,499.