Software Architecture for Multimodal User Input - FLUID

Abstract

Abstract. Traditional ways to handle user input in software are uncomfortable when an application wishes to use novel input devices. This is especially the case in gesture based user interfaces. In this paper we describe these problems and as a solution we present an architecture and an implementation of a user input toolkit. We show that the higher level processing of user input such as gesture recognition requires a whole new kind of paradigm. The system we designed and implemented -FLexible User Input Design (FLUID) -is a lightweight library that can be used in different kinds of software. The potential application areas include all system where novel input devices are in use: virtual reality, entertainment systems and embedded systems.

On Device Driver Architectures for Virtual Reality Toolkits

“…The FLUID toolkit (Ilmonen & Kontkanen, 2002) deals mainly with the implementation of data history processing for multimodal input. This toolkit emphasizes the use of histories for multimodal access and depicts various access patterns for histories.…”

Upponurkka: An Inexpensive Immersive Display for Public VR Installations

“…The software is based on the FLUID camera tracking code [1] and it tracks two color balls with two cameras. The tracking is performed with 30 Hz sampling frequency and it works very well in the viewing area which is about 3x3m in front of the screens.…”