Facebook’s facial recognition software is now as accurate as the human brain, but what now?

Facebook’s facial recognition research project, DeepFace (yes really), is now very nearly as accurate as the human brain. DeepFace can look at two photos, and irrespective of lighting or angle, can say with 97.25% accuracy whether the photos contain the same face. Humans can perform the same task with 97.53% accuracy. DeepFace is currently just a research project, but in the future it will likely be used to help with facial recognition on the Facebook website. It would also be irresponsible if we didn’t mention the true power of facial recognition, which Facebook is surely investigating: Tracking your face across the entirety of the web, and in real life, as you move from shop to shop, producing some very lucrative behavioral tracking data indeed.

The DeepFace software, developed by the Facebook AI research group in Menlo Park, California, is underpinned by an advanced deep learning neural network. A neural network, as you may already know, is a piece of software that simulates a (very basic) approximation of how real neurons work. Deep learning is one of many methods of performing machine learning; basically, it looks at a huge body of data (for example, human faces) and tries to develop a high-level abstraction (of a human face) by looking for recurring patterns (cheeks, eyebrow, etc). In this case, DeepFace consists of a bunch of neurons nine layers deep, and then a learning process that sees the creation of 120 million connections (synapses) between those neurons, based on a corpus of four million photos of faces.

Read more

Study Shows that Intensity of Facebook Use Can Be Predicted by Reward-related Activity in the Brain

Neuroscientists at Freie Universität Berlin show a link between reward activity in the brain due to discovering one has a good reputation and social media use

A person’s intensity of Facebook use can be predicted by activity in the nucleus accumbens, a reward-related area of the brain, according to a new study published by neuroscientists in the Languages of Emotion Cluster of Excellence at Freie Universität Berlin. Dr. Dar Meshi and his colleagues conducted this first ever study to relate brain activity (functional MRI) to social media use. The study was published in the latest issue of the open-access journal Frontiers in Human Neuroscience.

The researchers focused on the nucleus accumbens, a small but critical structure located deep in the center of the brain, because previous research has shown that rewards —including food, money, sex, and gains in reputation — are processed in this region.

“As human beings, we evolved to care about our reputation. In today’s world, one way we’re able to manage our reputation is by using social media websites like Facebook,” says Dar Meshi, lead author of the paper. Facebook is the world’s largest social media channel with 1.2 billion monthly active users. It was used in the study because interactions on the website are carried out in view of the user’s friends or public and can affect their reputation. For example, Facebook consists of users “liking” posted information. This approval is positive social feedback, and can be considered related to their reputation.

All 31 participants completed the Facebook Intensity Scale to determine how many friends each participant had, how many minutes they each spent on Facebook, and general thoughts. The participants were selected to vary widely in their Facebook Intensity Scale scores.

First, the subjects participated in a video interview. Next, the brain activity of the subjects was recorded, by using functional magnetic resonance imaging, in different situations. In the scanner, subjects were told whether people who supposedly viewed the video interview thought highly of them, and subjects also found out whether people thought highly of another person. They also performed a card task to win money.

Results showed that participants who received positive feedback about themselves produced stronger activation of the nucleus accumbens than when they saw the positive feedback that another person received. The strength of this difference corresponded to participants’ reported intensity of Facebook use. But the nucleus accumbens response to monetary reward did not predict Facebook use.

“Our study reveals that the processing of social gains in reputation in the left nucleus accumbens predicts the intensity of Facebook use across individuals,” says Meshi. “These findings expand upon our present knowledge of nucleus accumbens function as it relates to complex human behavior.”

Regarding the potential for social media addiction and the effects of social media on education quality, these results may provide important motivation for clinical research and for further research on learning. As Meshi says, “Our findings relating individual social media use to the individual response of the brain’s reward system may also be relevant for both educational and clinical research in the future.” The authors point out, however, that their results do not determine if positive social feedback drives people to interact on social media, or if sustained use of social media changes the way positive social feedback is processed by the brain.