Publication
Title
A multi-modal AI approach for intuitively instructable autonomous systems
Author
Abstract
We present a multi-modal AI framework to intuitively instruct and control Automated Guided Vehicles. We define a general multi-modal AI architecture, which has a loose coupling between three different AI modules, including spoken language understanding, visual perception and Reinforcement Learning navigation. We use the same multi-modal architecture for two different use cases implemented in two different platforms: an off-road vehicle, which can pick objects, and an indoor forklift that performs automated warehouse inventory. We show how the proposed architecture can be used for a wide range of tasks and can be implemented in different hardware, demonstrating a high degree of modularity.
Language
English
Source (journal)
International journal on advances in systems and measurements
Publication
2023
Volume/pages
16 :1&2 (2023) , p. 1-13
Full text (open access)
Full text (publisher's version - intranet only)
UAntwerpen
Faculty/Department
Research group
Project info
A Connected Brain-sized network – Design of a distributed connectivity layer for combining different heterogeneous deep learning systems.
Publication type
Subject
Affiliation
Publications with a UAntwerp address
External links
Record
Identifier
Creation 24.11.2023
Last edited 25.05.2024
To cite this reference