On Tuesday November 13th, Professor Linda Smith of Indiana University will give a talk on the topic of infant learning and development. The talk will be at 13:00 in Room B1.09 of the Computer Science Building. Linda runs the Cognitive Development lab in Bloomington Indiana.
Title: Learning from the infant’s point of view
How do infants learn their first words in a noisy environment? How do they progress from being slow incremental learners to rapid learners who appropriately generalize categories and concepts from minimal experience? How might the answers to these questions make for smarter, more nuanced, machine learning? We have used head cameras to collect egocentric views (and parent talk) in the home from the perspective of infants and toddlers (1 month olds to 30 month olds, with no experimenters present, 1000 hours of head camera video) and in a naturalistic toy room environment in the laboratory (about 200 hours of head-mounted eye tracking yielding both the ego-centric view and the gaze within that view). Our analyses indicate four principles we believe to be key to human prowess in visual category recognition: (1) Learn a massive amount about very few individual entities (and little bit about lots of other individual things); (2) Learn a massive amount about a very few categories (and a little bit about lots of other categories); (3) Learn about small selective sets at different points in time; (4) Self-generate the data for learning (with some help from mom and dad). The implications for both human cognition and machine learning will be discussed.