Visual Servoing with Geometrically Interpretable Neural Perception

TL;DR

This paper introduces a deep learning method for visual servoing that provides geometrically interpretable visual features, enabling easier and more flexible robotic control without explicit image processing.

Contribution

It presents a neural network trained with controller-guided supervision to produce interpretable visual features for visual servoing, enhancing flexibility and interpretability.

Findings

Validated on simulated and real robotic experiments.

Achieved geometrically interpretable visual features.

Demonstrated improved flexibility in visual servoing.

Abstract

An increasing number of nonspecialist robotic users demand easy-to-use machines. In the context of visual servoing, the removal of explicit image processing is becoming a trend, allowing an easy application of this technique. This work presents a deep learning approach for solving the perception problem within the visual servoing scheme. An artificial neural network is trained using the supervision coming from the knowledge of the controller and the visual features motion model. In this way, it is possible to give a geometrical interpretation to the estimated visual features, which can be used in the analytical law of the visual servoing. The approach keeps perception and control decoupled, conferring flexibility and interpretability on the whole framework. Simulated and real experiments with a robotic manipulator validate our approach.