Learning to Poke by Poking: Experiential Learning of Intuitive Physics

TL;DR

This paper presents a deep learning approach for a robot to learn intuitive physics through extensive poking interactions, enabling better manipulation by modeling dynamics directly from images.

Contribution

It introduces a joint deep neural network model for forward and inverse dynamics estimation from visual data, trained with real-world robotic poking experiences.

Findings

Joint modeling outperforms alternative methods.

Learning in feature space reduces pixel prediction complexity.

Over 400 hours of robotic interaction data used for training.

Abstract

We investigate an experiential learning paradigm for acquiring an internal model of intuitive physics. Our model is evaluated on a real-world robotic manipulation task that requires displacing objects to target locations by poking. The robot gathered over 400 hours of experience by executing more than 100K pokes on different objects. We propose a novel approach based on deep neural networks for modeling the dynamics of robot's interactions directly from images, by jointly estimating forward and inverse models of dynamics. The inverse model objective provides supervision to construct informative visual features, which the forward model can then predict and in turn regularize the feature space for the inverse model. The interplay between these two objectives creates useful, accurate models that can then be used for multi-step decision making. This formulation has the additional benefit that it is possible to learn forward models in an abstract feature space and thus alleviate the need of predicting pixels. Our experiments show that this joint modeling approach outperforms alternative methods.