Vision-based deep execution monitoring

TL;DR

This paper introduces a vision-based execution monitoring system for robots in uncharted environments, utilizing deep learning for object recognition, Bayesian methods for relation discovery, and reinforcement learning for visual search to improve task reliability.

Contribution

It presents an integrated approach combining deep learning, Bayesian inference, and reinforcement learning for robust visual execution monitoring in unstructured environments.

Findings

Effective object recognition with DCNNs.

Successful relation discovery using non-parametric Bayes.

Improved recovery from failures through learned visual search policies.

Abstract

Execution monitor of high-level robot actions can be effectively improved by visual monitoring the state of the world in terms of preconditions and postconditions that hold before and after the execution of an action. Furthermore a policy for searching where to look at, either for verifying the relations that specify the pre and postconditions or to refocus in case of a failure, can tremendously improve the robot execution in an uncharted environment. It is now possible to strongly rely on visual perception in order to make the assumption that the environment is observable, by the amazing results of deep learning. In this work we present visual execution monitoring for a robot executing tasks in an uncharted Lab environment. The execution monitor interacts with the environment via a visual stream that uses two DCNN for recognizing the objects the robot has to deal with and manipulate, and a non-parametric Bayes estimation to discover the relations out of the DCNN features. To recover from lack of focus and failures due to missed objects we resort to visual search policies via deep reinforcement learning.