Decoding Facial Recognition: A Revolution in Neuroscience
By Alyssa Britton
Take a moment; look around. Chances are you’ll see some familiar faces, whether they’re photos of celebrities on a magazine cover or the family, friends, or coworkers near you. How did you recognize them? You probably didn’t take the time to deliberately analyze each of their features, from the width of their noses to the tone of their skin to the angle of their eyebrows. Instead, without thinking, your brain instantly assessed each person’s features, recognized those features, and linked those perceptions to your memories of that person. Until now, the way your brain accomplishes such an extraordinary feat has been little more than speculation. However, thanks to recent research, we now know that to recognize faces, our brains employ different neurons to respond to each feature of the face, and the collective response of all the facial neurons generate recognition.1
In the 1960’s, neuroscientist Jerry Lettvin introduced his theory of memory and recognition that has since become the generally accepted model for facial recognition. Lettvin proposed that the brain designates a few neurons for every known concept; these neurons fire upon seeing a familiar entity, allowing one to recognize specific people or objects.2 He called these concept-specific neurons “grandmother cells” after the idea that there is a specific cell in each person’s brain that responds to and recognizes the sight or memory of his or her grandmother.2
Now, decades later, new studies are challenging Lettvin’s ideas. In a recent paper published in Cell, Le Chang and Doris Y. Tsao, researchers at the California Institute of Technology, claim to have cracked the code for facial recognition, calling into question Lettvin’s concept of grandmother cells. Their discovery marks “quite a revolution in neuroscience,” according to University of Leicester neuroscientist Professor Rodrigo Quian Quiroga.2
Chang and Tsao developed an algorithm that converts the characteristics of the human face to fifty numeric variables that measure both quantitative features (such as the width of a person’s nose) and qualitative ones (such as skin tone). By observing the individual responses of approximately 200 neurons in the region of the brain governing facial recognition in macaque monkeys, researchers linked the electrical impulses generated by the neurons, collectively called brainwaves, to each of the fifty variables.1
With the brainwaves of facial recognition decoded, researchers then began to encode them – that is, convert the primates’ brainwaves to digital images of faces. After showing the primates previously unseen faces, the researchers used each monkey’s brainwaves to reconstruct the faces with shocking accuracy.2 Chang and Tsao’s research “disavows the long-standing assumption that [grandmother cells] encode specific facial identities,” instead providing evidence that certain neurons in a wide network respond to specific facial characteristics.1
Chang and Tsao’s discoveries, however, raise a number of ethical questions about how their research might translate into actual technologies that can be used on humans. If forensic investigations begin to utilize this technology, then does decoding a person’s brain waves violate one’s Fifth Amendment ‘right to remain silent’? As neuroscientists further decode the brain–possibly unlocking the codes to memories, thoughts, or dreams–will the contents of a person’s mind be treated as physical property, subject to laws of search and seizure?
When asked about such concerns, Tsao said, “I can’t really relate to that. For me it’s great that we can understand something that’s been so mysterious.”2 As Tsao subtly points out, in the midst of such ethical contemplation, it can be all too easy to overlook the possible benefits of new discoveries. Perhaps instead of allowing speculation alone to inspire fear of Chang and Tsao’s groundbreaking discovery, we can look to the future of neuroscience with optimism and wonder.
- Chang, Le and Doris Y. Tsao. “The Code for Facial Identity in the Primate Brain.” Cell 169.6 (2017): 1013-1028. Web. 29 June 2017.
- Devlin, Hannah. “Scientists discover how the brain recognises faces – by reading monkeys’ minds.” The Guardian, 1 June 2017. Web. 29 June 2017.