Researchers were able successfully create a model of machine learning that can learn words by using footage recorded by a toddler who was wearing a headcam. The findings, which were published this week by Science, could shed new light on the ways children learn language and potentially inform researchers’ efforts to build future machine learning models that learn more like humans.
According to previous research, children begin to learn their first words between 6 and 9 months old. By the second birthday of a child, they have around 300 vocabulary words. The mechanics of how children learn to associate words with meaning remain unclear, and are a subject for scientific debate. Researchers from New York University’s Center for Data Science tried to explore this gray area further by creating an AI model that attempted to learn the same way a child does.
The researchers used over 60 hours worth of audio and video recordings from a head-mounted camera that was strapped on a toddler named Sam. The camera was worn by the toddler intermittently, starting at age…
