IntPhys: A Framework and Benchmark for Visual Intuitive Physics Reasoning
arXiv:1803.07616
Abstract
In order to reach human performance on complexvisual tasks, artificial systems need to incorporate a sig-nificant amount of understanding of the world in termsof macroscopic objects, movements, forces, etc. Inspiredby work on intuitive physics in infants, we propose anevaluation benchmark which diagnoses how much a givensystem understands about physics by testing whether itcan tell apart well matched videos of possible versusimpossible events constructed with a game engine. Thetest requires systems to compute a physical plausibilityscore over an entire video. It is free of bias and cantest a range of basic physical reasoning concepts. Wethen describe two Deep Neural Networks systems aimedat learning intuitive physics in an unsupervised way,using only physically possible videos. The systems aretrained with a future semantic mask prediction objectiveand tested on the possible versus impossible discrimi-nation task. The analysis of their results compared tohuman data gives novel insights in the potentials andlimitations of next frame prediction architectures.
References in corpus (15)
- Adam: A Method for Stochastic Optimization
- Conditional Generative Adversarial Nets
- Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
- Learning to Segment Object Candidates
- A Novel Performance Evaluation Methodology for Single-Target Trackers
- Interaction Networks for Learning about Objects, Relations and Physics
- Semi-Supervised Learning with Context-Conditional Generative Adversarial Networks
- A Disentangled Recognition and Nonlinear Dynamics Model for Unsupervised Learning
- Learning Physical Intuition of Block Towers by Example
- Visual Interaction Networks
- To Fall Or Not To Fall: A Visual Approach to Physical Stability Prediction
- Learning A Physical Long-term Predictor
- A Comparative Evaluation of Approximate Probabilistic Simulation and Deep Neural Networks as Accounts of Human Physical Scene Understanding
- 3D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstruction
- Generalizable Features From Unsupervised Learning
Cited by in corpus (18)
- Flexible Neural Representation for Physics Prediction
- ThreeDWorld: A Platform for Interactive Multi-Modal Physical Simulation
- CoPhy: Counterfactual Learning of Physical Dynamics
- Baby Intuitions Benchmark (BIB): Discerning the goals, preferences, and actions of others
- CATER: A diagnostic dataset for Compositional Actions and TEmporal Reasoning
- Physion: Evaluating Physical Prediction from Vision in Humans and Machines
- Unsupervised Discovery of 3D Physical Objects from Video
- Occlusion resistant learning of intuitive physics from videos
- The Perceptimatic English Benchmark for Speech Perception Models
- Forward Prediction for Physical Reasoning
- A Benchmark for Modeling Violation-of-Expectation in Physical Reasoning Across Event Categories
- Unsupervised Intuitive Physics from Past Experiences
- Learning Generalizable Physical Dynamics of 3D Rigid Objects
- AVoE: A Synthetic 3D Dataset on Understanding Violation of Expectation for Artificial Cognition
- AGENT: A Benchmark for Core Psychological Reasoning
- Predicting the Physical Dynamics of Unseen 3D Objects
- PHYRE: A New Benchmark for Physical Reasoning
- Curiosity-driven Intuitive Physics Learning