Awesome

Intuitive Physics Reading List

This is an awesome reading list for intuitive physics raning from cognitive studies to computational studies. The project aims to facilitate knowledge sharing in the community of intuitive physics and build a startup library for beginners. The list will be continuously updated.

Researchers
Survey
Understand Physical Commonsense
Predict Physical Events
Intervene in the Physical Environment

Researchers

Josh Tenenbaum, MIT, Computational Cognitive Science Group (CoCoSci Group).
Kevin Smith, MIT, Computational Cognitive Science Group (CoCoSci Group).
Peter Battaglia, DeepMind.
Jiajun Wu, Stanford, Stanford Vision and Learning Lab (SVL).
Tomer Ullman, Harvard, Computation, Cognition, and Development Lab (CoCoDev)
Judith Fan, UCSD->Stanford, Cognitive Tools Lab
Todd Gureckis, NYU, Computation & Cognition Lab
Ernest Davis, NYU.
Yixin Zhu, PKU, Cognitive Reasoning Lab (CoRe)
Ilker Yildirim, Yale, Cognitive & Neural Computation Lab (CNCL)
Joseph Lim, USC->KAIST, Cognitive Learning for Vision and Robotics Lab (CLVR)
Kenneth D. Forbus, Northwestern University, Qualitative Reasoning Group (QRG)

Survey

Intuitive Physics: Current Research and Controversies, TiCS 2018. Hongjing Lu's review on intuitive physics.
Dark, Beyond Deep: A Paradigm Shift to Cognitive AI with Humanlike Common Sense (Section 4), Engineering 2020. An interpretation of intuitve physics from a CV perspective.
A Survey on Machine Learning Approaches for Modelling Intuitive Physics, arXiv 2202.06481. A review on the approaches for modeling intuitive physics.

Understand Physical Commonsense

Object permanence in 3 1/2-and 4 1/2-month-old infants, Psychol 1987. A psycological study to test object permanence in infants using the VoE paradigm (look longer at the impossible events).
Intuitive Physics Learning in a Deep-learning Model Inspired by Developmental Psychology, Nature Human Behaviour 2022. A computational study that introduces a deep model to learn physical concepts in the VoE paradigm.
X-VoE: Measuring eXplanatory Violation of Expectation in Physical Events, ICCV 2023. A comprehensive benchmark to assess AI's grasp of intuitive physics with explanatory capacities.

Predict Physical Events

Simulation as an Engine of Physical Scene Understanding, PNAS 2013. The first attempt to computationally verify intuitive physics engine with simulation.
Functional Neuroanatomy of Intuitive Physical Inference, PNAS 2016. A piece of evidence for the functional part of intuitive physics engine in human brain.
Learning Physical Intuition of Block Towers by Example, ICML 2016. A learning approach using CNN to predict whether and how the block towers will fall.
Interaction Networks for Learning about Objects, Relations and Physics, NeurIPS 2016. The first physical dynamics prediction approach with relation modeling between physical objects.
Neural Foundations of Mental Simulation: Future Prediction of Latent Representations on Dynamic Scenes, NeurIPS 2023. An AI+Neuroscience study discussing different modeling methods in physical prediction with behavior and neural dynamic evidence.

Intervene in the Physical Environment

Understanding Tools: Task-Oriented Object Modeling, Learning and Recognition, CVPR 2015. A framework to understand tool use by learning essential physical concepts from one rational RGB-D video.
PHYRE: A New Benchmark for Physical Reasoning, NeurIPS 2019. A physical reasoning benchmark studying how to reach a goal by placing balls in the scenes.
Rapid Trial-and-error Learning with Simulation Supports Flexible Tool Use and Physical Reasoning, PNAS 2020. A physical reasoning benchmark similar to PHYRE studying how to reach a goal by placing various tools in the scenes.
On the Learning Mechanisms in Physical Reasoning, NeurIPS 2022. A meta-analysis of the learning mechanisms in solving PHYRE-like physical reasoning problems.
I-PHYRE: Interactive Physical Reasoning, ICLR 2024. A challenging benchmark to address real-time interactivity in physical reasoning.