ArmAIF: A practical introduction of active inference for arm perception and action

Keywords

Loading...
Thumbnail Image

Issue Date

2021-02-19

Language

en

Document type

Journal Title

Journal ISSN

Volume Title

Publisher

Title

ISSN

Volume

Issue

Startpage

Endpage

DOI

Abstract

This paper is a tutorial on active inference using high dimensional inputs, such as raw images, providing a simpli ed theory and code for the understand of the algorithm of arm perception and action. The results are test cases in which the algorithms performance is shown in multiple di erent situations, for example when the sensory input is limited to just vision or propriocep- tion. In the case of vision, for instance, overlap of the internal and sensory state is crucial. thus far the algorithm armAIF has performed as expected and is capable of using active inference for minimizing variational free energy.

Description

Citation

Supervisor

Faculty

Faculteit der Sociale Wetenschappen