Augmenting Physical Simulators with Stochastic Neural Networks: Case Study of Planar Pushing and Bouncing

Abstract

An efficient, generalizable physical simulator with universal uncertainty estimates has wide applications in robot state estimation, planning, and control. In this paper, we build such a simulator for two scenarios, planar pushing and ball bouncing, by augmenting an analytical rigid-body simulator with a neural network that learns to model uncertainty as residuals. Combining symbolic, deterministic simulators with learnable, stochastic neural nets provides us with expressiveness, efficiency, and generalizability simultaneously. Our model outperforms both purely analytical and purely learned simulators consistently on real, standard benchmarks. Compared with methods that model uncertainty using Gaussian processes, our model runs much faster, generalizes better to new object shapes, and is able to characterize the complex distribution of object trajectories.

arXiv commentIROS 2018

References in corpus (5)

• Auto-Encoding Variational Bayes
• Visual Dynamics: Probabilistic Future Frame Synthesis via Cross Convolutional Networks
• Stochastic Variational Video Prediction
• A Fast Stochastic Contact Model for Planar Pushing and Grasping: Theory and Experimental Validation
• GP-SUM. Gaussian Processes Filtering of non-Gaussian Beliefs

Cited by in corpus (6)

• Residual Policy Learning
• Data-Augmented Contact Model for Rigid Body Simulation
• 3D-OES: Viewpoint-Invariant Object-Factorized Environment Simulators
• Residual Reinforcement Learning from Demonstrations
• Hybrid modeling: Applications in real-time diagnosis
• Modular meta-learning in abstract graph networks for combinatorial generalization