Vanderbilt
Register
Front Page
Links
Search
To search for more information on this topic, choose from one of the following services:
Research pinpoints brain activity
Experts use facial recognition system also to identify cars, birds
Gauthier by David F. Salisbury
A team of psychologists from Vanderbilt and Yale universities have shown that the area of the brain that allows people to pick out quickly faces in a crowd also helps birdwatchers identify different species of birds with a glance and car experts distinguish between different makes and models of automobiles.In a study reported in the February issue of the journal Nature Neuroscience, the researchers used the powerful technique of functional magnetic resonance imaging to compare the regions of the brain that veteran birdwatchers and car fanciers use to recognize pictures of human faces, birds and automobiles.
They found that a small region located on the bottom of the brain, called the fusiform face area(FFA), activates when birdwatchers view human faces and birds. Similarly, the face area becomes active when auto experts look at human faces and automobiles.
The face area is part of the visual cortex. It was identified as the site within the brain that specializes in facial recognition almost 10 years ago. For a number of years scientists have known that people recognize other human faces in a manner that differs fundamentally from the way in which they identify other objects in their environment. They tend to immediately identify faces as individuals, while they tend to identify objects initially by class. People also have a much harder time identifying upside-down faces than they do other kinds of objects viewed from similarly unfamiliar viewpoints.
“One possibility...is that it may be possible to design a training program that could significantly improve autistic children’s ability to recognize other people.”
—Isabel Gauthier, Assistant professor of psychology
“We think that the method we use for identifying objects is by focusing on individual features,” said Isabel Gauthier, the assistant professor of psychology at Vanderbilt who was the first author on the study. “But the method we use for identifying faces is a holistic one. That is, we process the entire facial image at the same time.”
Gauthier continued, “We have a lot of experience with faces and a particular kind of experience with faces. But do we really have a separate system for facial recognition? Or do we have a general recognition system that has been extensively trained to identify faces?”
To answer this question, Gauthier and her colleagues — Pawel Skudlarski, John Gore and Adam Anderson at the Yale University Medical School — recruited eight bird experts and 11 car experts. They put each person in an fMRI machine and then asked them to look at photos of faces, birds and cars. The fMRI allowed the researchers to identify the areas in each person’s brain that became active as he or she processed the various images.
The fact that the face area became active when birdwatchers were identifying birds and auto experts were identifying automobiles indicates that this high-powered recognition system is not limited to identifying faces but is used for other purposes as well. In fact, the study revealed that the most knowledgeable experts were those who relied on the FFA most extensively.
The study also proved that the facial recognition system is not limited to certain types of objects. That was demonstrated when it worked equally as well with cars and birds. “The system does not care what the objects look like,” Gauthier added.
The common denominator between facial recognition and FFA-mediated expert recognition appears to be an extensive training process, Gauthier said. She and Michael J. Tarr of Brown University established this association in a previously published study using a class of objects that they invented called “Greebles.” Sixty Greebles are divided up into two genders and five families. Like faces, the Greebles all have a similar appearance and differ in relatively subtle ways.
The researchers then paid a group of students to become Greeble experts. They found that it only took a couple of hour-long training periods before the students were able to identify all the Greebles correctly. But it took them nearly 10 sessions before the face-area-based recognition system kicked in and the students began identifying individual Greebles as individuals as rapidly as they could identify them by gender or family.
In addition to providing new insights into how the brain is organized, these studies may also provide new ideas for the treatment of autism and similar conditions. In an article scheduled to appear in the April issue of the Archives of General Psychiatry, Gauthier — working with Robert T. Schultz and colleagues at Yale — shows that young adults with autism do not rely on the face-area recognition system to discriminate between faces as much as their peers. Instead, they tend to rely on another part of the brain associated with the identification of objects. That helps explain why individuals with autism have difficulty identifying faces.
“One possibility suggested by our work is that it may be possible to design a training program that could significantly improve autistic children’s ability to recognize other people by increasing their use of the facial recognition system,” Gauthier said.
This research was funded by the National Institutes of Health.
