What do babies see? What are they looking at? The answers to these questions are very different for younger babies than they are for older infants, children, and adults. Characterized by some high-contrast edges in simple patterns, these early scenes also contain the very materials needed to build a strong foundation for human vision.
That’s the finding of a new study, “An edge-simplicity bias in the visual input of young infants,” published May 10 in Advances in Science by IU researchers Erin Anderson, Rowan Candy, Jason Gold and Linda Smith.
The starting assumption for everyone thinking about the role of experience in visual development has always been that on the scale of everyday experience, visual input is pretty much the same for everyone. However, this study says, no, visual input changes with development. It’s not the same for everyone. Contribution to daily life for very young infants appears to be unique at this age.”
Linda Smith, principal investigator, professor in the Department of Psychological and Brain Sciences
Previous studies in the laboratory and clinic have shown that young infants prefer to see simple, high-contrast scenes with large black stripes and a checkerboard. The current study is the first to ask to what extent these preferences are input to their daily lives. “To see what young babies see and see,” says Anderson, a former postdoctoral researcher in Smith’s Cognitive Development Lab, she and her colleagues put head cameras on infants to wear at home during daily life activities.
“You can buy ‘baby flash cards’ for newborns that show these simple, high-contrast images,” she explains. “What the head camera videos show, what this work shows, is that young infants find these kinds of images all around them in their everyday lives, just by looking at things like lights and the corners of the ceiling.”
“What we found is a very specific, early ‘diet’ for visual development,” adds Smith. “As with food, young infants don’t start out with rich, complex meals or pizza, but with simple, developmentally specific food.”
Previous work has recognized the critical nature of this early period for the future development of human vision. For example, infants born with visual abnormalities such as cataracts or those in orphanages with limited visual experiences have been shown to have lifelong visual impairments. The current study offers some preliminary data to address these shortcomings. It also has important implications for building visual AI systems, which also acquire stronger visual skills when training begins with the same simple, high-contrast visual content.
“The sheer scale of the influx of everyday life”
To identify the properties of visual input in infants aged approximately three to 13 months, researchers attached head-mounted video cameras to 10 infants and their 10 adult caregivers, collecting and analyzing 70 hours of visual documentation of daily life at home. Clear differences emerge between infant and adult image contents with a greater concentration of simple patterns and high-contrast edges in infants’ than adults’ views.
Smith concludes that the reason for these views is not just that infants will turn their heads to see features of the world they can see, but that parents or caregivers are likely to put them in places they like. to look at things. “You have to think about why they’re where they are. There’s probably some implicit natural knowledge on the part of parents to leave infants where they want to look at things. Mom won’t bother you if you don’t fuss,” she observes.
However, is this small group of participants from Bloomington, Indiana, representative of infants more broadly around the world? To answer this question, Smith’s lab conducted the same experiment with a collaborator in a small, bustling fishing village in Chennai, India where electricity is scarce and much of daily life happens outdoors. And while the 6-month-old and 12-month-old head-cam images looked very different from their Bloomington counterparts, the younger infants share a common diet of high-contrast edges and simple patterns in both Chennai and Bloomington.
Bigger pictures, past and future
Smith and her colleagues have also shown that the same sequence of images improves the training of visual AI systems. In a follow-up to the current study, published in the 2023 Neural Information Processing Systems Conference Proceedings, they found that if you train an AI system by first feeding it images characteristic of early infancy, it has more success learning to recognize visual images than if you feed it random images developmental series or simply provide images typical of an adult’s daily life. The most accurate developmental sequence produced the best results.
Their work opens new avenues for evolutionary speculation. As Smith explains, “One of the things I’ve always asked as a graduate student—and maybe we’ll have a chance to answer that—is why human babies have such slow motor development. They spend about three months just hearing and seeing for another six months with little posture and head control Why do horses come out so slowly?
This research suggests that “over evolutionary time these slow, gradual and optimized biases work to create a very intelligent visual and auditory system,” he says. “This is a story that could be told.”
Meanwhile, their work raises new questions about the visual content of early infancy and its role in the developing visual system, whether human or AI.
Other researchers include IU Bloomington professors Rowan Candy in the School of Optometry and Jason Gold in the Department of Psychological and Brain Sciences.
Source:
Journal Reference:
Anderson, EM, et al. (2024) A simplicity bias in visual input in young infants. Advances in Science. doi.org/10.1126/sciadv.adj8571.
