Jacob Boesen Madsen

Temporal Coherence Strategies for Augmented Reality Labeling

Temporal coherence of annotations is an important factor in augmented reality user interfaces and for information visualization. In this paper, we empirically evaluate four different techniques for annotation. Based on these findings, we follow up with subjective evaluations in a second experiment. Results show that presenting annotations in object space or image space leads to a significant difference in task performance. Furthermore, there is a significant interaction between rendering space and update frequency of annotations. Participants improve significantly in locating annotations, when annotations are presented in object space, and view management update rate is limited. In a follow-up experiment, participants appear to be more satisfied with limited update rate in comparison to a continuous update rate of the view management system.

Aspects of what makes or breaks a museum AR experience

The paper critically evaluates central aspects of an iPad AR application developed for a museum context. The application is designed for children aged 8 to 12 and mixes AR and mini-game elements to convey dramatized historical events. The game has been deployed for roughly 3 months and the findings in the paper are supported by extensive in-application activity logging. Actual usage of the application at the museum proved far less extensive than envisaged. Hypotheses for this finding are presented and discussed, with support from the logging data.

Handheld Visual Representation of a Castle Chapel Ruin

We have experienced rapid development in Augmented Reality (AR) systems and platforms in the world of cultural heritage, namely in cultural settings and historical museums. However, we still face a range of challenges to design an AR system that meets the requirements for an AR installation working autonomously in a cultural heritage setting for an extended duration. This article describes the development of two installations for the visualization of a 3D reconstruction of a castle chapel, running autonomously during open hours in the location of a castle museum. We present a convincing 3D visualization running at interactive frame-rates on modern tablets. In one installation, the tablet is connected to a large screen TV for an immersive experience, and, in another, the tablet is hand-held, thus facilitating translational freedom in the chapel. Both installations allow unsupervised usage during museum visiting hours. Based on in-field observations and on-device logging of application usage, user behavior is analyzed and evaluated. Results indicate that users spent a limited amount of time using the application and did not fully explore the visual area of the chapel. In order for the user to spend more time with the application, additional information must be presented to the user.

An interactive visualization of the past using a situated simulation approach

This paper describes aspects of the development of an interactive installation for visualizing a 3D reconstruction of a historical church chapel in Kolding, Denmark. We focus on three aspects inherent to a mobile Augmented Reality development context; 1) A procedure for combating gyroscope drift on handheld devices, 2) achieving realistic lighting computation on a mobile platform at interactive frame-rates and 3) an approach to relocation within this applications situated location without position tracking. We present a solution to each of these three aspects. The development is targeted a specific application, but the presented solutions should be relevant to researchers and developers facing similar issues in other contexts. We furthermore present initial findings from everyday usage by visitors at the museum, and explore how these findings can be useful in connection with novel technology for facilitating information transfer to a museum audience. The installation is in active commercial use and is currently logging further user interactions via in-application logging for future investigations in line with this project.

Differences in human audio localization performance between a HRTF- and a non-HRTF audio system

Spatial audio solutions have been around for a long time in real-time applications, but yielding spatial cues that more closely simulate real life accuracy has been a computational issue, and has often been solved by hardware solutions. This has long been a restriction, but now with more powerful computers this is becoming a lesser and lesser concern and software solutions are now applicable. Most current virtual environment applications do not take advantage of these implementations of accurate spatial cues, however. This paper compares a common implementation of spatial audio and a head-related transfer function (HRTF) system implementation in a study in relation to precision, speed and navigational performance in localizing audio sources in a virtual environment. We found that a system using HRTFs is significantly better at all three performance tasks than a system using panning.