TU Berlin

Quality and Usability Labkolloquium_WS09/10

Inhalt des Dokuments

zur Navigation

Forschungskolloquium WS 09/10

Mondays, starting October 19, 2009
TEL Auditorium 1 and 2 (20th floor), Ernst-Reuter-Platz 7, 10587 Berlin

This research colloquium is a weekly event with various invited speakers. It is open to anyone who is interested in the general area of usability and human-computer interaction. Researchers in this area will present overviews of their work. The colloquium is organized by Deutsche Telekom Laboratories. If you have any questions, please contact Shiva Sundaram or Hamed Ketabdar or Sebastian Möller

Please contact us if you want your email to be added to our colloquium mailing list.

Colloquium Talks
TITLE: Prototyping research tools. (POSTPONED)

SPEAKER: Nicolas Villar, Microsoft Research Cambridge.

The scope of my work is the development of novel user interface technologies and new interaction techniques that improve the experience of working and playing with technology. This area of research benefits from an applied approach, whereby novel devices and concepts for interaction are realized to the degree that they can be placed in the hands of users to be evaluated. The practice of prototyping is central to this process: the degree to which we can effectively iterate on an idea, and refine a concept towards a working prototype, has a direct impact on our abilities to effectively explore a given problem area, identify important research questions, and ultimately garner insights through deployment and user evaluation. In this talk I want to discuss some of the current challenges involved prototyping hardware-oriented interactive devices, and present my efforts in developing a platform for rapid development of high-fidelity prototypes. Dragonfly is a modular hardware platform - based on the .NET Micro Framework -which provides support for three key areas in making fully functional devices: electronic circuit assembly, software development and form-factor design.HOST: Michael Rohs.

TITLE: Overcoming the Smallness of Small Devices. (POSTPONED)

SPEAKER: Sean Gustafson, HPI, Potsdam.

The inherent smallness of mobile phones and other personal devices limits their ability to accept input and provide output, preventing the rich interaction available on conventionally sized devices. In this talk, I will discuss methods of exploiting our innate ability to think spatially and how this can be used to effectively “enlarge” small devices. Along with a survey of current research in this area, I will present my work on visualizing off-screen content, my study into the foundation of tilt-based interaction for mobile devices, and other ongoing research in our lab. These projects are examples of a growing research trend that attempts to regain some of the bandwidth lost when devices shrunk to their current size.

HOST: Michael Rohs
TITLE: Automatic Processes in the Experience of Design

SPEAKERS: Prof. Dr. Ad Pruyn; Dr. Thomas van Rompay
Department of Marketing Communication & Consumer Psychology
University of Twente, The Netherlands


Research indicates that the traditional perspective on consumer behavior based on conscious information processing is in need of revision. Many choices, evaluations and product perceptions are strongly affected by subtle environmental or contextual influences that elude conscious awareness.

After introducing the importance of such processes with respect to consumer behavior in general, two automatic processes (processing fluency & embodied perspective taking) in the context of design experience will be discussed and experimental evidence will be presented.

HOST: Anna E. Pohlmeyer, Stefan Schaffer.

ROOM: "AUDITORIUM" Floor 20, Ernst Reuter Platz-7, Berlin.

TIME: 18.15 - 20.00 hrs.

TITLE: Cooking an Ontology-based Spoken Dialogue System: a proposal
SPEAKER: Joana Paulo Pardal

The main goal of this thesis is to study how different levels of knowledge stored in ontologies can be used to facilitate the creation of new coaching dialogue systems capable of domain reasoning. We take McGuinness' ontologies spectrum - and related work - to support the division of a cooking ontology into increasingly complex knowledge levels.The hypothesis being tested is whether ontologies can be used to enrich a coaching spoken dialogue system and be used in it in such way that the system can abstract the source of domain-specific knowledge - related to the tasks being coached - focusing only on the dialogue phenomena. The integration of ontological knowledge should be done with few architecture adaptions to the dialogue system so that when adding a new domain - a new class of tasks - minor changes in special modules are sufficient. Software engineering techniques for systems integration, like extension-based modularization and design patterns, are to be considered and extensively used to minimize the number of architecture adaptions.

HOST: Dmitry Butenkov

Harald Gründl


Ancient rituals, images and myths are still part of today’s consumer culture. Although they appear contemporary transformed, they still stabilize groups of people, help them to overcome social crises and divert them from other cultural groups and articulations. The lecture of the designer and design theorist Harald Gruendl ranges from the passage rite of fashion to paradigm change to sustainable design, from the fork to the city using transitional design strategies.

HOST: Corinna Schmidt
TITLE: Does mimicry of co-speech gestures result from alignment in mental representations?

SPEAKER: Lisette Mol, Communication and Cognition group, Faculty of Humanities, Tilburg University, The Netherlands.

For certain co-speech hand gestures, instances where the same encoding patterns were repeated across speakers have been found (i.e. de Fornel, 1992; Tabensky, 2001; Kimbara, 2006). The question I want to address in the first part of this talk is whether this mimicry of hand gestures is best seen as automated and direct behavioral mimicry following the perception-behavior link (Chartrand & Bargh, 1999), or whether it is highly dependent on meaning and thus more likely to result from alignment in mental representations (Pickering & Garrod, 2004). In two studies we have shown that certain iconic gestures were mimicked only if their meaning matched the co-occurring speech. This suggests that alignment in iconic gestures may indeed result from alignment in representations, much like for example lexical alignment.
In the second part of my talk, I will present some newer research that looks at alignment in gesture in more detail. In a task in which a participant and a confederate took turns describing short routes, the confederate systematically varied the hand shape of her path gestures as well as the perspective from which they were produced (either in the horizontal or vertical plane), while the co-occurring speech was kept constant. The results of this study provide further evidence that similarities between the confederate’s and the participant’s hand gestures resulted from alignment in mental representations, rather than from direct behavioral mimicry. In a follow up study, we have taken a closer look at how consistency of perspective in speech and gesture influences alignment in gesture (and speech).

HOST: Christine Kühnel.

SPEAKER: Prof. Dr. Friedrich Wolfram Heubach

HOST: Sandra Buchmüller
TITLE: Pattern orientierte Entwicklung von Sprachschnittstellen.

SPEAKER: Dirk Schnelle-Walka, Technische Universität Darmstadt.

Voice User Interface Design wird heutzutage noch immer als eine Kunst betrachtet, die sich nicht ohne Weiteres mit der Herangehensweise eines Software Ingenieurs vereinbaren lässt. So wollen bisherige Ansätze hauptsächlich die Konservierung von Best-Practice Know-How in Form von Guidelines erreichen. Meistens beschränken sich diese jedoch auf eine Darstellung wie man es besser nicht machen sollte. In diesem Vortrag wird eine Herangehensweise vorgestellt, wie Voice User Interface Design unter Verwendung von Design Patterns sinnvoll umgesetzt werden kann.
Hierdurch werden Methoden, die sich bereits im Software Engineering Bereich erfolgreich durchgesetzt haben, auf den Bereich des Voice User Interface Designs übertragen.

Please note the special time and date: 18.15-20.00 hrs, Auditorium 3

Issues in aesthetics and usability of interactive systems

Noam Tractinsky

HOST: prometei, Robert Schleicher
TITLE: Living Security - Concepts for Continuous Security Management of Service Oriented Systems.

SPEAKER: Ruth Breu, University of Innsbruck, AUT.

The new generation of IT systems is characterised by interconnectedness and the possibility to conduct business processes across enterprises and systems. Arising scenarios e.g. in health care and traffic management have demonstrated that the new technologie+G1s around service orientation have the potential to create new business models and new forms of
cooperation. On the other side cooperation of independent stakeholders and systems comes along with increased requirements to security of data and processes. In addition service oriented systems are in most cases dynamically evolving. In this challenging context this talk presents
an approach to Living Security. Living Security is targeted towards continuous security management from compliance and design of IT services to systems operation.

HOST: Jean-Pierre SEIFERT.
Please notice the special time and date: 18.15-20.00 hrs, Auditorium 3

Experience Design: Erleben gestalten

SPEAKER: Prof. Dr. Marc Hassenzahl

"Produkte müssen funktionieren". Das ist eine der Grundüberzeugungen der Mensch-Technik-Interaktion. Dabei betont besonders die Usability die optimale Passung eines Produkts zur Physiologie und den kognitiven Möglichkeiten seiner Benutzer. Typische Gestaltungsziele sind Gesundheitsförderlichkeit, Gebrauchstauglichkeit und Performanz. "Nutzererleben" (user experience) geht einen Schritt weiter. Es fordert die Passung eines Produkts zu grundlegenden menschlichen Bedürfnissen, wie sozialer Nähe, Kompetenz, Autonomie, oder Stimulation. Ziel ist es Wohlbefinden und nicht so sehr mehr Leistung erzeugen. Der Unterschied ist ein wichtiger: Während Ergonomie und Usability ihre Wurzeln in Arbeitsorganisationen haben, deren vorrangiges Ziel es ist, die Effektivität und Effizienz ihrer Mitarbeiter zu sichern und zu verbessern, betont Nutzererleben den Menschen und die grundlegenden Motivationen seines Handelns. Effektivität und Effizienz ist so gesehen kein Ziel der Menschen, sondern eines der Organisationen für die sie arbeiten. Eine auf Nutzererleben konzentrierte Gestaltung betont den persönlichen Nutzen, also beispielsweise die positiven Kompetenzerlebnisse, die sich aus Effektivität und Effizienz subjektiv ergeben können, und versucht ihn durch Gestaltung zu verstärken und zu verbessern.

Der Vortrag führt in den Ansatz der "user experience" ein und diskutiert seine Auswirkungen auf Produktevaluation und -gestaltung anhand konkreter Beispiele.

HOST: Julia Seebode, Stefan Schaffer

TITLE: On Services: Culture Engines, Habit Traps and Fields of Flesh

Benedict Singelton

Service design is more talked about than practiced, and the talk is a mixture of hands-on savvy with big-hearted polemics: methodological ‘toolkits’ and ruminations on ‘design thinking’ converge with - rather sketchy - promises of ‘community’, ‘sustainability’, and ‘wellbeing’. The four-year research programme I’ll be talking about takes a very different position on a field whose wide and rapid emergence is as surprising as the speed with which it has generated its own orthodoxy.  
If a service is an act that benefits someone else, its design is an attempt to give human co-operation a particular shape: service design hacks the social in an effort to choreograph it. Services produce roles, habits, expectations, opportunities, etiquette, body postures, wealth, poverty, and political positions. Design's role in this production is not a priori ‘good’ nor ‘bad’, but complex, inevitable, powerfully charged, always questionable and never to be rushed into.  
This research locates an alternative practice of service design at the intersection of historical attempts to push design ‘beyond the object‘ with recent forays into new political terrain stirred up across a range of disciplines. It attempts to unlock a role for service design as a medium of experimentation with new institutional forms. In the course of this talk, we will cross paths with the relationship between distributed organisations and all-out war, puzzle over whether a pregnant woman can be said to provide a service to her unborn child, outline tendencies of commercial design research we must strictly speaking call 'criminal' and 'surreal', think about how biotech challenges how we understand what ‘working for a living’ means and how we understand human habits... that is, in general, consider means of (re)starting conversation about what service design is, does, might be and could do.      

Katharina Bredise

TITLE: The People are the network.

SPEAKER: Dr. Pan Hui, Deutsche Telekom Laboratories (INET-Group).

The people are the network. People form networks in different ways, for example through physical proximity, mobility,and social connectivity. Equipped with modern mobile communication devices, they become part of the communication network itself. Via long range cellular channels and short range local radio communication, information propagate ubiquitously throughout the human network. Mobile networks have dual properties; it is a social network, and at the same time also a communication network. Understanding human network characteristics is key to designing better
communication systems, improve urban planning, and even stop the spread of epidemics. My research focus is on humans and humanity: understanding human networks and improving humanity in the digital world, which includes to promote freedom of speech, reduce media monopoly, minimize unwanted information, and improve the connectivity of rural and developing area.

HOST: Hamed Ketabdar and Shiva Sundaram.

Please notice: the talk will take place in room 1118/19, 11th floor, Ernst-Reuter-Platz 7

“Surveillance Image Quality for Recognition Tasks: Definitions, Testing, Standardization, and Current Trends”.
SPEAKER: Prof. Mikołaj LESZCZUK. Department of Telecommunications Krakow, Poland


Monitoring of public safety (traffic, intersections, mass events, stations, airports and other public ur-ban areas) using the transmission and analysis of video content gains in the recent period on the im-portance of the overall increase in crime and acts of terrorism (attacks on the WTC, public transport in London and Madrid). Quality of Experience (QoE) of video content used for entertainment (digital TV, including HDTV, and multimedia on the Internet) differs materially from the QoE of surveillance video used for recognition tasks in CCTV monitoring, because in the latter case, the subjective satis-faction of the consignee shall recede in achieving the given function (event detection, object recogni-tion). What’s more, the quality of surveillance video, acceptable from a subjective point of view (using the operator’s monitoring), is largely separate from the objective video quality useful in computer processing (Computer Vision).
Assessment principles, and especially maximization of the surveillance video quality, are a relatively new field. State-of-the-Art solutions were limited mainly to optimizing the network QoS parameters (bandwidth, packet loss probability), eventually, for surveillance video, classical methods (created for entertainment content), like the PSNR (Peak Signal English-to-Noise Ratio) or SSIM (Structural SI-Milarity) were applied. Issues of quality measurements for surveillance video are partly reflected in the few, quite outdated standards and recommendations (including ITU-T P.910, “Subjective video quality assessment methods for multimedia applications”, 1999 and ITU-T P.912: “Subjective video quality assessment methods for recognition tasks”, 1998), mainly introducing basic definitions, me-thods of testing and psychophysical experiments (Multiple Choice Method, Single Answer Method, Timed Task Method), the distinction between the Real-Time Viewing and Viewer-Controlled View-ing, the concept of the scenarios.
The most influential body in the surveillance video quality is VQiPS (Video Quality in Public Safety) Workgroup, funded by the U.S. Department of Homeland Security, set up in 2008. So far VQiPS achievements include coordination of the various organizations whose goal is to create standards for surveillance video. The current work focuses on: education of users and the development of specifica-tions for surveillance video quality. It has been observed that even different surveillance video applica-tions can combine common elements that affect image quality specification. As a consequence, VQiPS creates a set of use cases that are independent of the application, at the same time enriching them with instructions for users to adapt the VQiPS specific standards to their own applications. VQiPS will also create a consistent terminology of concepts related to the quality of video utility and related equip-ment.
Video surveillance is an issue seen by numerous bodies aimed at protecting citizens against “perma-nent surveillance” in Orwellian style. Among these, we should mention a Liberty Group (dedicated to human rights), an Open Europe organization, the Electronic Frontier Foundation or the Ethics Board of the INDECT project. This matter was also one of the main themes (“Citizens Security Needs Ver-sus Citizens Integrity”), of the Fourth Security Conference Research Conference organized by the European Commission (September 2009). In practice, alternative methods of at least partial protection of privacy do exist, based on the selective only monitoring of figureheads, automatic erase of fac-es/license plates not related with the investigation or data hiding techniques using digital watermark-ing.

HOST: Sebastian Moeller.
Please notice: the talk will take place in room 1118/19, 11th floor, Ernst-Reuter-Platz 7

Language Acquisition Embedded into Tutor-Robot Interaction

SPEAKER: Dr. Martin Heckmann, Honda Research Institute Europe

Children acquire language to a large extend in the interaction with their caregivers. Inspired by this observation we develop computational models and artifacts for the acquisition of language in an interactive scenario. The artifact bootstraps its representations with little a priori knowledge and can be taught by a human tutor. In this presentation I will highlight different aspects of our research in this domain.

First I want to talk about the learning of speech features as well as word and sub-word units. As speech features we explore a set of hierarchical spectro-temporal features which are learned in an unsupervised fashion based on the observed speech data. I will demonstrate that these spectro-temporal features carry information complementary to purely spectral features and that this information is beneficial for speech recognition.
Furthermore, I will show how incremental word learning can be bootstrapped via the use of phone-like speech units emerging from an unsupervised clustering process and efficient alignment of a few training samples. Next I will introduce an auditory attention system enabling a robot to filter relevant from irrelevant acoustic events.

Finally, I will briefly highlight a system which learns associations between acoustic labels and visual representations in interaction with its tutor. The auditory processing is based on the building blocks described above and the whole system is implemented on our humanoid robot Asimo.

HOST: Sebastian Moeller
Please notice: the talk will take place in room 1118/19, 11th floor, Ernst-Reuter-Platz 7

Expressive Sound Synthesis for Animation

SPEAKER: Cécile Picard

The main objective is to provide tools for an expressive and real-time synthesis of sounds resulting from physical interactions of various objects in a 3D virtual environment. Indeed, these sounds, such as collisions sounds or sounds from continuous interaction between surfaces, are difficult to create in a pre-production process since they vary drastically depending on the interaction type and objects. To achieve this goal, two approaches are proposed; the first one is based on simulation of physical phenomena responsible for sound production, the second one is based on the processing of a recordings database.

Short Bio:
Cécile Picard obtained her Ph.D. in Computer Science at INRIA, France, in the REVES team in December 2009. Her Ph.D. research focused on real-time sound rendering for virtual reality.
After receiving her diploma in Mechanical Engineering with specialization in Acoustics in 2005 at the Université de Technologie de Compiègne (UTC), France, she continued in the domain of sound and obtained in 2006 a Msc. in Sound and Vibrations at the Chalmers University in Goteborg, Sweden.
Her current research focuses on a usable tool to allow soundtrack composers, sound designers and electroacoustic music composers to create auditory experiences (AudioGarden: http ://blog.audiogarden.org/pages/about).  In collaboration with Christian Frisson from the UCL (Be), they propose a novel approach that combines physically based synthesis methods and handling of prerecordings into an adequate multimodal user interface for sound composition and design.

HOST: Alexander Müller
This talk will be 30-40 minutes

TITLE: Student Project: Evaluation of the Joy of Use Button 

HOST: Robert Schleicher
TITLE: Analyzing user interaction with web-based geospatial systems

SPEAKER: Jens Ingensand

In recent years web-based geospatial systems have been created for different purposes such as way-finding systems, online atlases and collaborative systems.
In order to accommodate the fact that an increasing number of people have access to geospatial data and technology, research has begun to focus on the impact and the utilization of such systems. However to the best of our knowledge no formal research framework currently exists that allows for analyzing the interaction of real-world users with such systems and that takes into account the spatial nature of these systems. Furthermore few studies regarding the usability of web-based geospatial systems have been conducted and most of these studies were set up in artificial lab-environments with few mostly poorly described test-users.
We have conducted three case studies where real-world users interact with three web-based geospatial systems. The important findings of those studies were that there are noticeable gender differences regarding spatial navigation, that user satisfaction appears to be biased by the user's previous experience with geographic information technologies and that the type of the input device used (e.g. mouse or touchpad) has a considerable influence on task-completion time and user satisfaction. The above results indicate that the user's demographic parameters and background have a significant effect on the user's spatial interaction strategies, performance and satisfaction.
To further fuel research on usability aspects of web-based geospatial systems, we provide the necessary tools and methods for conducting remote- and laboratory based testing of such systems.

Born: 13.10.1977
Nationality: German
Current position: Lecturer and assistant at the GIS-lab, Swiss Federal Institute of Technology (EPFL), Lausanne, Switzerland
MSc in Applied Informatics With a Major in Interaction Design, Chalmers, University of Technology, Gothenburg, Sweden MSc in Human and Economic Geography, University of Gothenburg, Sweden Dr. és sciences (PhD) Swiss Federal Institute of Technology (EPFL), Lausanne, Switzerland 

HOST: Christine Kühnel

Amir Mehmood

Hamed Ketabdar and Shiva Sundaram
TITLE: Current Trends in Speech Translation

SPEAKER: Florian Metze

HOST: Tim Polzehl
TITLE: Current Trends in Speech Translation

SPEAKER: Florian Metze

HOST: Tim Polzehl

SPEAKER: Lennard J. Davis

HOST: Tom Bieling



Schnellnavigation zur Seite über Nummerneingabe