Intuitive physics learning in a deep-learning model inspired by developmental psychology
Luis S. Piloto, Ari Weinstein, Peter Battaglia et al. · Nature Human Behaviour · 2022
'Intuitive physics' enables our pragmatic engagement with the physical world and forms a key component of 'common sense' aspects of thought. Current artificial intelligence systems pale in their understanding of intuitive physics, in comparison to even very young children. Here we address this gap between humans and machines by drawing on the field of developmental psychology. First, we introduce and open-source a machine-learning dataset designed to evaluate conceptual understanding of intuitive physics, adopting the violation-of-expectation (VoE) paradigm from developmental psychology. Second, we build a deep-learning system that learns intuitive physics directly from visual data, inspired by studies of visual cognition in children. We demonstrate that our model can learn a diverse set of physical concepts, which depends critically on object-level representations, consistent with findings from developmental psychology. We consider the implications of these results both for AI and for research on human cognition.
Cited by 118 works. doi:10.1038/s41562-022-01394-8 Open-access PDF
Authors
References
- Long Short-Term Memory
- Building machines that learn and think like people
- Relational inductive biases, deep learning, and graph networks
- Improved protein structure prediction using potentials from deep learning
- Addition and subtraction by human infants
- Origins of knowledge.
- Object permanence in five-month-old infants
- Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm
- A simple neural network module for relational reasoning
- Object permanence in 3½- and 4½-month-old infants.