A Compositional Object-Based Approach to Learning Physical DynamicsDownload PDF

Published: 21 Jul 2022, Last Modified: 22 Oct 2023ICLR 2017 PosterReaders: Everyone
Abstract: We present the Neural Physics Engine (NPE), a framework for learning simulators of intuitive physics that naturally generalize across variable object count and different scene configurations. We propose a factorization of a physical scene into composable object-based representations and a neural network architecture whose compositional structure factorizes object dynamics into pairwise interactions. Like a symbolic physics engine, the NPE is endowed with generic notions of objects and their interactions; realized as a neural network, it can be trained via stochastic gradient descent to adapt to specific object properties and dynamics of different worlds. We evaluate the efficacy of our approach on simple rigid body dynamics in two-dimensional worlds. By comparing to less structured architectures, we show that the NPE's compositional representation of the structure in physical interactions improves its ability to predict movement, generalize across variable object count and different scene configurations, and infer latent properties of objects such as mass.
TL;DR: We propose a factorization of a physical scene into composable object-based representations and also a model architecture whose compositional structure factorizes object dynamics into pairwise interactions.
Conflicts: mit.edu
Keywords: Deep learning, Unsupervised Learning
Community Implementations: [![CatalyzeX](/images/catalyzex_icon.svg) 1 code implementation](https://www.catalyzex.com/paper/arxiv:1612.00341/code)
19 Replies

Loading