Projects

We present Adaptive Pointing, a novel approach to addressing the common problem of accuracy when using absolute pointing devices for distant interaction. The intention behind this approach is to improve pointing performance for absolute input devices by implicitly adapting the Control-Display gain to the current user's needs without violating users' mental model of absolute-device operation. First evaluation results show that Adaptive Pointing leads to a significant improvement compared with absolute pointing in terms of movement time (19%), error rate (63%), and user satisfaction.

In contrast to the pioneers of multimodal interaction e.g. Richard Bolt in the late seventies, today s researchers can benefit of a wide variety of existing interaction techniques, devices and frameworks. Although these tools are available, the usage of them is still a great challenge particularly in terms of usability. A major issue results from the trade-off between the functionality of the system and the simplicity of use. We introduce a novel visual user interface concept which is especially designed to ease the design and development of post-WIMP user interfaces including multimodal interaction. It provides an integrated design environment for our interaction library Squidy based on high-level visual data flow programming combined with zoomable user interface concepts. The user interface offers a simple visual language and a collection of ready-to-use devices, filters and interaction techniques. We specifically address the trade-off between functionality and simplicity by utilizing the concept of semantic zooming which enables dynamic access to more advanced functionality on demand. Thus, developers as well as interaction designers are able to adjust the complexity of the Squidy user interface to their current need and knowledge.

We present Adaptive Pointing, a novel approach to addressing the common problem of accuracy when using absolute pointing devices for distant interaction. First, we discuss extensively some related work concerning the problem-domain of pointing accuracy when using absolute or relative pointing devices. As a result, we introduce a novel classification scheme to more clearly discriminate between different approaches. Second, the Adaptive Pointing technique is presented and described in detail. The intention behind this approach is to improve pointing performance for absolute input devices by implicitly adapting the Control-Display gain to the current user‟s needs without violating users‟ mental model of absolute-device operation. Third, we present an experiment comparing Adaptive Pointing with pure absolute pointing using a laser-pointer as an example of an absolute device. The results show that Adaptive Pointing results in a significant improvement compared with absolute pointing in terms of movement time (19%), error rate (63%), and user satisfaction.

This project will demonstrate a new approach to employing users gaze in the context of human-computer interaction. This new approach uses gaze passively in order to improve the speed and precision of manually controlled pointing techniques. Designing such gazeaugmented manual techniques requires an understanding of the principles that govern the coordination of hand and eye. This coordination is influenced by situational parameters (task complexity, input device used, etc.), which this project will explore in controlled experiments.

dc.title:

dc.contributor.author: Rädle, Roman; König, Werner A.; Reiterer, Harald

In this paper we provide a threefold contribution to the Surface Computing (SC) community. Firstly, we will discuss frameworks such as Reality-based Interaction which provide a deeper theoretical understanding of SC. Secondly, we will introduce our ZOIL user interface paradigm for SC on mobile, tabletop or wall-sized devices. Thirdly, we will describe our two software tools Squidy and ZOIL UI Framework which have supported and facilitated our iterative design of SC prototypes.

dc.title:

dc.contributor.author: Schmidt, Toni; König, Werner A.; Reiterer, Harald

Previous studies have argued for the use of gaze-assisted pointing techniques (MAGIC) in improving human-computer interaction. Here, we present experimental findings that were drawn from human performance of two tasks on a wall-sized display. Our results show that a crude adoption of MAGIC across a range of complex tasks does not increase pointing performance. More importantly, a detailed analysis of user behavior revealed several issues that were previously ignored (such as, interference of corrective saccades, increased decision time due to variability of precision, errors due to eye-hand asynchrony, and interference with search behavior) which should influence the development of gaze-assisted technology.

In this paper we present our experiences with longitudinal study designs for input device evaluation. In this domain, analyzing learning is currently the main reason for applying longitudinal designs. We will shortly discuss related research questions and outline two case studies in which we used different approaches to address this issue. Finally, we will point out future research tasks in the context of longitudinal evaluation methods.

With growing quantity and complexity of available information, novel user interface concepts are needed which support users natural seeking behavior and provide natural ways of interaction. We discuss general user requirements and present design principles which we identified as crucial for the successful design of a user-centered visual information-seeking system. We also illustrate novel input modalities and display techniques such as multi-touch tables, gesture tracking systems and physical tokens which lead towards a more reality-based and thus natural interaction.

We employ Laserpointer-Interaction as an intuitive, direct and flexible interaction concept particularly for large, highresolution displays which enforce physical navigation. We therefore demonstrate general applicability and suitability by applying Laserpointer-Interaction to a wide range of domains from scientific applications and formal experiments to artistic installations for the broad public.

Human beings perceive their surroundings based on sensory information from diverse channels. However, for humancomputer interaction we mostly restrict the user on visual perception. To investigate the effect of additional tactile feedback on pointing and selection tasks we conducted a comparative evaluation study based on ISO 9241-9. The 20 participants performed horizontal and vertical one-directional tapping tasks with hand gesture input with and without tactile feedback on a large, high resolution display. In contrast to previous research we cannot confirm a benefit of tactile feedback on user performance. Our results show no significant effect in terms of effective index of performance and even a significant higher error rate for horizontal target alignment when using tactile feedback. Furthermore we found a significant difference in favor of the horizontal target alignment compared to the vertical one in terms of the effective index of performance.

In this paper we introduce the novel user interface paradigm ZOIL (Zoomable Object-Oriented Information Landscape). ZOIL is aimed at unifying all types of local and remote information items with their connected functionality and with their mutual relations in a single visual workspace as a replacement of today s desktop metaphor. This workspace can serve as an integrated work environment for traditional personal information management (PIM), but can also be used for PIM tasks in a wider sense. By formulating ZOIL s fundamental design principles we describe the interaction style, visualization techniques and interface physics of a ZOIL user interface. Furthermore we discuss ZOIL s ability to provide nomadic PIM environments for mobile and stationary use.

Aufgrund des limitierten Sehvermögens des Menschen muss sich ein Anwender vor großen, hochauflösenden Displays frei bewegen können, um kleinste Details pixelgenau wahr zu nehmen oder einen Überblick über die gesamte Darstellungsfläche von mehreren Quadratmetern zu erhalten. Im Gegensatz zu konventionellen Eingabegeräten wie Maus und Tastatur beschränken Laserpointer den Anwender nicht in seiner Bewegungsfreiheit, sondern ermöglichen unabhängig von der Position zum Display eine sehr intuitive und direkte Art der Interaktion. In diesem Beitrag wird eine Interaktionsbibliothek vorgestellt, welche im Hinblick auf Präzision und verzögerungsfreier Steuerung erstmals auch den Einsatz von Laserpointer-Tracking bei großen, hochauflösenden Displays ermöglicht. In einem Vergleichsexperiment wurde die Interaktionsbibliothek in Kombination mit einem Infrarotlaserpointer gegenüber einer klassischen Maus als Standardeingabegerät bei unterschiedlichen Distanzen evaluiert. Der signifikante Performancevorteil der Maus von 12,5% scheint in Anbetracht der gewonnenen Bewegungsfreiheit und der unmittelbaren Interaktionsweise mit dem Laserpointer eher gering ins Gewicht zu fallen. Im Vergleich zu bisherigen Systemen konnte der Rückstand des Laserpointers um über 50% reduziert werden, was größtenteils auf die geringe Bewegungslatenz, das präzise Tracking und die wirksame Kompensation des Zitterns der Hand zurückzuführen ist. Ferner wurde bei der Studie ein signifikanter Distanzeffekt beim Laserpointer hinsichtlich Performance und Fehlerrate festgestellt.

Since large, high-resolution displays (LHRD) are capable of visualizing a large amount of very detailed information, users have to move around in front of these displays to gain either in-depth knowledge or an overview. However, conventional input devices such as mouse and keyboard restrict users mobility by requiring a stable surface on which to operate. We present a flexible and intuitive interaction technique based on an infrared laserpointer, a technique that allows identical use from any point and distance. In particular, our laserpointer interaction satisfies the demands of LHRD in the areas of mobility, accuracy, and interaction speed. The solution presented is technically designed as a generic interaction library whose flexibility and general applicability was verified by using it on two very different systems a planar 221" Powerwall and a curved 360° panoramic display. Furthermore, a comparative evaluation study with 16 participants was conducted on the Powerwall to compare the performances of a conventional mouse and our laserpointer by means of a unidirectional tapping test at varying distances (ISO 9241-9). The statistically significant performance advantage of the mouse (13%) appears marginal considering the intuitive and direct mode of interaction in using the laserpointer and the flexibility gained by its use, both of which are fundamental requirements for LHRDs. In comparison to previous systems and evaluations, we were able to reduce the laserpointer s performance lag by over 50%. This result is achieved mainly by our precise tracking method, the minimized delay, and the effective jitter compensation.