Modeling Expectation Violation in Intuitive Physics with Coarse Probabilistic Object Representations
Kevin Smith · Lingjie Mei · Shunyu Yao · Jiajun Wu · Elizabeth Spelke · Josh Tenenbaum · Tomer Ullman

Thu Dec 12th 10:45 AM -- 12:45 PM @ East Exhibition Hall B + C #177

From infancy, humans have expectations about how objects will move and interact. Even young children expect objects not to move through one another, teleport, or disappear. They are surprised by mismatches between physical expectations and perceptual observations, even in unfamiliar scenes with completely novel objects. A model that exhibits human-like understanding of physics should be similarly surprised, and adjust its beliefs accordingly. We propose ADEPT, a model that uses a coarse (approximate geometry) object-centric representation for dynamic 3D scene understanding. Inference integrates deep recognition networks, extended probabilistic physical simulation, and particle filtering for forming predictions and expectations across occlusion. We also present a new test set for measuring violations of physical expectations, using a range of scenarios derived from developmental psychology. We systematically compare ADEPT, baseline models, and human expectations on this test set. ADEPT outperforms standard network architectures in discriminating physically implausible scenes, and often performs this discrimination at the same level as people.

Author Information

Kevin Smith (MIT)
Lingjie Mei (MIT)
Shunyu Yao (Princeton University)
Jiajun Wu (Google)
Elizabeth Spelke (Harvard University)

Elizabeth Spelke teaches at Harvard University, where she is Professor of Psychology and Co-Director of the Mind, Brain, and Behavior Initiative. She studies the origins and nature of knowledge of objects, persons, space, and number, by assessing behavior and brain function in human infants, children, human adults and non-human animals. She is a member of the National Academy of Sciences and the American Academy of Arts and Sciences, and was cited by Time Magazine as one of America's Best in Science and Medicine, her honors include the Distinguished Scientific Contribution Award of the American Psychological Association and the William James Award of the American Psychological Society.

Josh Tenenbaum (MIT)

Josh Tenenbaum is an Associate Professor of Computational Cognitive Science at MIT in the Department of Brain and Cognitive Sciences and the Computer Science and Artificial Intelligence Laboratory (CSAIL). He received his PhD from MIT in 1999, and was an Assistant Professor at Stanford University from 1999 to 2002. He studies learning and inference in humans and machines, with the twin goals of understanding human intelligence in computational terms and bringing computers closer to human capacities. He focuses on problems of inductive generalization from limited data -- learning concepts and word meanings, inferring causal relations or goals -- and learning abstract knowledge that supports these inductive leaps in the form of probabilistic generative models or 'intuitive theories'. He has also developed several novel machine learning methods inspired by human learning and perception, most notably Isomap, an approach to unsupervised learning of nonlinear manifolds in high-dimensional data. He has been Associate Editor for the journal Cognitive Science, has been active on program committees for the CogSci and NIPS conferences, and has co-organized a number of workshops, tutorials and summer schools in human and machine learning. Several of his papers have received outstanding paper awards or best student paper awards at the IEEE Computer Vision and Pattern Recognition (CVPR), NIPS, and Cognitive Science conferences. He is the recipient of the New Investigator Award from the Society for Mathematical Psychology (2005), the Early Investigator Award from the Society of Experimental Psychologists (2007), and the Distinguished Scientific Award for Early Career Contribution to Psychology (in the area of cognition and human learning) from the American Psychological Association (2008).

Tomer Ullman (Harvard)

More from the Same Authors