Inhalt des Dokuments
Forschungskolloquium 2011
| Date: | Mondays, starting January 26, 2011 | |
|---|---|---|
| Time: | 14:15-15:45 | |
| Location: | 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 Hamed Ketabdar, Sebastian Möller or Klaus-Peter Engelbrecht.
Please contact us if you want your email to be added to our colloquium mailing list.
For an archive of past colloquium talks please check Forschungskolloquium 2010.
| DATE | TALK DESCRIPTION |
|---|---|
| 23.04.2012 | Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 1 SPEAKER: Poitr Majdak, Acoustics Research Institute der Öster. Akademie der Wissenschaften, Wien HOST: Sascha Spors Last updated 19.01.2012 by Hamed Ketabdar |
| 20.02.2012 through 09.04.2012 | No Colloquium (Holiday)! |
| 13.02.2012 | Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 1 TITLE: Adaptation of Aesthetic Appreciation SPEAKER: Claus-Christian Carbon BRIEF DESCRIPTION: Personal taste develops over time and is highly susceptible for Zeitgeist-dependent effects. Using an adaptation paradigm often used in the domain of face research (e.g., Carbon & Ditye, 2011), we could show in a series of experiments that not only the representation of designs (e.g., car designs, Carbon, 2010) or artworks (Carbon & Leder, 2006), but also taste quickly adapts towards adaptors (Carbon, Ditye, & Leder, 2006). The present talk will not only demonstrate adaptation towards specific design characteristics in product design, but even so for specific art-specific properties, thus underlining the power of adaptation for the development and advancement of taste in important domains of our everyday life. BIO: CCC is head of the Department of General Psychology and Methodology at the University of Bamberg, in Bamberg / Germany. His research concentrates on fields of empirical aesthetics, design appreciation, face recognition/face processing/prosopagnosia (face blindness), optical illusions, cognitive maps, advancement of methods and on applied cognition (design; the role of innovation; HCI and ergonomics). HOST: Klaus-Peter Engelbrecht |
| 06.02.2012 | Location: Ernst-Reuter-Patz 7, 20th Floor, Auditorium 1 TITLE: Designing Gestural Interaction for Different Contexts SPEAKER: Tanja Döring BRIEF DESCRIPTION: Gestural interaction offers rich opportunities for human-computer interaction and has evolved to an important research topic in interaction design. Currently, different types of gesture sets are developed, and among these are, e.g., touch gestures, mobile phone gestures or in-air gestures. An important factor for designing gestures is the context, in which they are used. This talk will focus on designing gestural interaction for different contexts and present case studies from automotive and entertainment user interfaces. Different aspects of gestural interaction like gestures in combination with speech or tangible interaction will be discussed. SHORT BIO: Tanja Döring is a researcher and PhD student in the Digital Media Group at the University of Bremen. Her research interests include tangible, gestural and mobile interaction with interactive surfaces. From April 2008 until October 2011, Tanja worked in the "Pervasive Computer and User Interface Engineering Group" at the University of Duisburg-Essen. She has a master’s degree in computer science from the University of Hamburg. dm.tzi.de/de/people/staff/tanja-doering/ HOST: Katrin Wolf Last updated 20.01.2012 by Hamed Ketabdar |
| 30.01.2012 | Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 1 TITLE: Introduction to W3C Multimodal Interaction Working Group SPEAKER: Ingmar Kliche BRIEF DESCRIPTION: Almost 10 years ago the “Multimodal Interaction Working Group” was launched within the W3C to “extend the Web to allow users to dynamically select the most appropriate mode of interaction for their current needs”. This presentation will give an overview of the technologies developed within the Multimodal Working Group (such as Multimodal Architecture and Interfaces, InkML, EMMA, EmotionML, ...) and related developments in other W3C Working Groups (such as VoiceXML, SCXML, ...). BIO: Ingmar Kliche is a project manager at Telekom Innovation Laboratories (T-Labs). He received his Master of Science degree in Electrical Engineering from the Berlin University of Technology in 1997. He has been working in speech technology for more than ten years in various capacities, including implementation of voice and multimodal applications, speech technology consulting and R&D project management. Ingmar is a member of several W3C working groups, including the W3C Voice Browser Working Group and the W3C Multimodal Interaction Working Group. HOST: Ina Wechsung Last updated 11.01.2012 by Hamed Ketabdar |
| 23.01.2012 | Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 1 TITLE: Grasp Sensing for Human-Computer Interaction SPEAKER: Raphael Wimmer BRIEF DESCRIPTION: The ability to grasp objects was a prerequisite to man's use of hand tools. These tools - such as hammers, computer mice, or mobile phones - afford different ways of grasping them. Sensing and interpreting these grasps gives us additional information about a user's goals or requirements while using the tool. This talk presents advantages and applications of grasp sensing on mobile devices as well as challenges for sensing and interpreting grasps. SHORT BIO: Raphael Wimmer is a researcher in the Media Informatics Group at the University of Regensburg and a PhD student in the Media Informatics Group at the University of Munich. Raphael's research focuses on touch sensing technologies and novel interaction techniques. He developed CapToolKit, an open-source toolkit for prototyping capacitive sensing systems, and currently investigates how to enhance touch-sensing on small and deformable objects. Blog: raphaelwimmer.wordpress.com Publications: www.medien.ifi.lmu.de/team/raphael.wimmer/ HOST: Katrin Wolf |
| 19.01.2012 12:00-13:00 | Please notice exceptional time: Thursday 12:00 -13:00 h! Location: Ernst-Reuter-Platz 7, 18th Floor, Room Plicht SPEAKER: Konrad Krenzlin TITLE:Optimizing Auditory Stimuli for a Novel Brain-Computer-Interface Paradigm BRIEF DESCRIPTION: Brain-Computer-Interfaces (BCI) enable touchless and motionless interaction by classifying the brain's response to stimuli presented to the user. Usually, visual stimuli are used, while auditory stimuli are seldom. An exception is the PASS2D paradigm, which uses different artificial tones driving a text entry system. The choice of stimuli is very important not only for the performance of the classifier, but also for the usability of the BCI. This talk shows my work on using short natural phonemes with different characteristic instead of artificial tones as stimuli and presents promising results. SHORT BIO: Konrad Krenzlin is currently finishing his Bachelor in Technische Informatik at TU Berlin. Recently, he started a Master course in Audio Communication and Technology, also at TU Berlin. The talk is a presentation of his Bachelor thesis. HOST: Sebastian Möller |
| 16.01.2012 | Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 1 TITLE: Is phonetics still needed in speech synthesis research? SPEAKER: Petra Wagner BRIEF DESCRIPTION: Recent developments have shown a rapid decrease in genuinely phonetic topics in the field of speech synthesis research. This development can be traced back to several reasons:
BIO: Petra Wagner is full professor for Phonetics and Phonology at Bielefeld University. After her studies in Linguistic, she finished her dissertation about prediction and perception of German stress pattern in 2002 at Bonn University. Her research is concerned with perceptual prominence and its relation to linguistic entities; speech, language and gestural rhythm, other topics in prosody and speech synthesis. HOST: Benjamin Weiss Last updated 11.1.2012 by Hamed Ketabdar |
| 09.01.2012 | Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 1 TITLE: Laughing, crying and other non-verbal vocalisations in conversational speech SPEAKER: Jürgen Trouvain BRIEF DESCRIPTION: Affect in spoken communication can be identified based on prosodic features like average pitch, pitch variation, intensity, timing or voice quality. However, additional to speech, there are vocalisations like crying or laughing, prominently signaling affect. Analysing such vocalisations is important for, e.g., affective speech synthesis, improved speech recognition. SHORT BIO: Jürgen Trouvain is senior researcher at the Saarbrücken University, Computational Linguistics and Phonetics Institute. His 1995 masters’ thesis was about building and evaluating spectral and durational features of a German formant synthesis. He published his dissertation in 2004, entitled “Tempo Variation in Speech Production. Implications for Speech Synthesis”. Currently, he is working on laughter and other non-verbal vocalisations, non-native speech, Luxembourgish, von Kempelen's speaking machine. +++ There will be an ADDITIONAL SHORT PRESENTATON of Jürgen Trouvain's work on performance of humans in understanding ultra-fast speech synthesis. +++ HOST: Benjamin Weiss |
| 02.01.2012 | Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 1 TITLE: Information Retrieval using context-free Descriptions of Audio SPEAKER: Carola Trahms BRIEF DESCRIPTION: Freely accessible media provided in the Internet increase at a rapid rate. This media need to be indexed reasonably, to allow a fast and efficient search. Indexing media, such as audio clips, is by no means a trivial task. On the one hand, the indexes should be as precise and unambiguous as possible and, on the other hand, they should enable a most efficient query. Indexing media by their content is a very precise method and was successfully done in the past, but queries performed on media will not always search for the content, but for a certain association that is connected to it. Associations are a less exact method for describing media as they usually are context free and differ from person to person. The descriptions used in this thesis are spontaneous comments on audio clips given by a number of listeners. Finding a way to index media by such spontaneous comments is a challenging task. In this thesis, a method for clustering audio clips using context-free descriptions will be suggested. SHORT BIO: Carola Trahms studies Computer Engineering at TU Berlin. The talk is a presentation of her Bachelor thesis which she did at T-Labs. She has started a Master course in Computer Engineering at TU Berlin this year and works in T-Labs as a student worker. HOST: Shiva Sundaram |
| 19.12.2011 | Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 1 TITLE: Binaural Model-Based Speech Intelligibility Enhancement and Assessment in Hearing Aids SPEAKER: Anton Schlessinger BRIEF DESCRIPTION: In Europe, about one fifth of the population has difficulties with understanding speech in noisy and complex environments. Improving speech intelligibility in these conditions allows for the reintegration of the hearing impaired into an communication-oriented society, and restores individual well-being to a high degree. Commercially available hearing aid solutions are generally based on the amplification principle and successfully enhance speech understanding for severe hearing loss in silence. However, current hearing aid solutions do not reconstruct speech intelligibility in noisy surroundings to an extent that is required by the majority of the hearing impaired. Successful solutions that reconstruct the intelligibility of noise-corrupted speech are based on the principle of spatial sampling. By such means, a target speaker signal can be enhanced and the interference can be suppressed. In this talk, a set of standard binaural speech processors, that draw upon models of the auditory scene analysis, are reviewed, optimized and compared. These binaural speech processors are, furthermore, applied at the output of hearing aids with and without beamforming. As a result, two efficient spatial sampling schemes are combined to achieve high improvement of speech intelligibility in noisy environments. A statistical study on binaural parameters in different acoustic real-world scenes is given. Binaural parameters of the fine-structure of the waveform are compared with binaural parameters of the corresponding envelope. In addition, natural binaural parameters are compared to binaural parameters at the output of a set of hearing aids. As a result, the study provides a comprehensive insight into the behaviour of binaural parameters in noise, thereby sizing the possibilities of a binaural parameter-based source segregation. Furthermore, a stochastic optimization of binaural speech processors at the output of different front-ends as well as in different acoustic environments is performed. To that end, a genetic algorithm is applied, which maximizes an objective function of binaural speech intelligibility. The robustness of the optimized binaural speech processors is assessed throughout modified acoustic scenes. The holistic approach of a model-based improvement and a model-based assessment of speech intelligibility offers an efficient and task-oriented means for the improvement of speech intelligibility.However, the development of an objective function of binaurally and nonlinearly processed speech, still remains an unsolved problem. To date, there exists no comprehensive model of speech intelligibility. Therefore the talk concludes with a summary on recent advances towards such a model. BIO: Anton Schlesinger enrolled in the study of Media Technology at Ilmenau University of Technology, Germany, in 2000. In 2004/05 he took a one year break from his study for an internship as a consultant for classical room acoustics for recording studios and venues at the Walters-Storyk Design Group in Basel, Switzerland. Subsequently, he has been engaged as a founding member of a start-up for location-based services in Weimar, Germany. In August 2006, he received his Diplom in in Media Technology with his thesis work on the three-dimensional measurement and holographic reconstruction of sound fields. At the end of 2006, he joined the Laboratory of Acoustical Imaging and Sound Control at the Delft University of Technology, The Netherlands. In the succeeding years, he continued research in room acoustical holography and conducted his work in audiology that resulted in his PhD thesis ``binaural model-based speech intelligibility enhancement and assessment in hearing aids.'' Since March 2011, Anton Schlesinger is employed at the Institute of Communication Acoustics at the Ruhr-Universiät Bochum, Germany, as a postdoctoral researcher. His current scientific interest is in binaural models of speech intelligibility, classification and model-based speech intelligibility enhancement. HOST: Pablo Ramirez Last updated on 23.11.2011 by Hamed Ketabdar |
| 12.12.2011 | Please notice: There are two talks for this date! Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 1 First Talk: TITLE: Auswirkungen von verzögertem Systemfeedback im mobilen Kontext SPEAKER: Arne Denneler BRIEF DESCRIPTION: Berührungsempfindliche Bildschirme (Touchscreens) finden vor allem bei modernen mobilen Geräten große Verbreitung. Durch ihren Einsatz in Mobiltelefonen mit erweiterter Computerfunktionalität (Smartphones) gestatten sie es dem Nutzer, die unterschiedlichen Anwendungen intuitiv durch Anklicken oder Verschieben von grafischen Elementen zu bedienen. Der Verzicht auf eine reale Tastatur bringt allerdings einen entscheidenden Nachteil mit sich. Herkömmliche Tastaturen geben dem Nutzer beim Drücken einer Taste unmittelbar eine taktile und eine dezente auditive Rückmeldung (Feedback). Dieses Feedback liefert eine virtuelle Tastatur zunächst nicht. Deshalb werden bei der Verwendung virtueller Tastaturen meist künstliche Feedbacks erzeugt, die allerdings systembedingten Verzögerungen unterliegen. In einer vorangegangenen Studie wurde mit einem statischen Versuchsaufbau untersucht, wie sich unterschiedliche Verzögerungen bei künstlich erzeugtem Feedback auf das Eingabeergebnis und auf die Beurteilung durch die Nutzer auswirken. Im Rahmen der vorgestellten Bachelorarbeit wird die vorherige Untersuchung in Hinsicht auf mobile Situationen, also auf einen realistischeren Nutzungskontext, ausgeweitet und die Ergebnisse beider Studien miteinander verglichen. HOST: Julia Seebode Second Talk: TITLE: The effect of feature alteration in anthropomorphic robot heads on social facilitation SPEAKER: Patrick Ehrenbrink BRIEF DESCRIPTION: Human performance can be influenced by the presence of other persons. As artificial intelligence and automation develop, robots become an ever more important part of society. The presented study examines if social facilitation effects can be triggered by robots. Three robots that differ in human-likeness serve as artificial experimenters in a speed-task. Their influence on the participants' performance is investigated. BIO: Patrick Ehrenbrink is a student research assistant at the Quality and Usability Lab in Berlin. He is currently a Master's student of Human Factors at the Technische Universität Berlin. HOST: Ina Wechsung Last updated 8.12.2011 by Klaus-Peter Engelbrecht |
| 05.12.2011 | Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 1 TITLE: Building multimodal systems is for everyone: a html+voice tutorial SPEAKER: David Griol and Zoraida Callejas Carríon BRIEF DESCRIPTION: X+V (xhtml + voice) is a markup language for developing multimodal web-based applications. X+V is specially useful to bring voice to WWW applications, as it provides the flexibility and facilities of mature web technologies such as XTHML and XML Events, along with the possibility to encode voice interaction using a subset of the VoiceXML standard. We will present the basics of X+V and explain how to structure the multimodal interaction in several sample applications. Please bring your computers – it might get interactive. HOST: Ina Wechsung Last updated 24.11.2011 by Hamed Ketabdar |
| 28.11.2011 | Please notice: There are two talks for this date! Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 1 First talk: TITLE: Definition and Implementation of a Human Readable Storage and Interchange Format for Spatial Audio Scenes SPEAKER: Lukas Kaser BRIEF DESCRIPTION: Research in the domain of spatial audio created a paradigm: object-based audio. The separation of an audio source and its parameters, such as its positions in space, allows the specification of a format for the storage and interchange of spatial audio scenes. This thesis extends the Audio Scene Description Format (ASDF), which was proposed in literature, with the possibility to store time-varying spatial audio parameters. A flexible timing concept adapted from the Synchronized Multimedia Integration Language (SMIL) is specified and a solution for the notation of parameterized trajectories, defining the movement of sources, is proposed. This concepts ease the text-based authoring of complex audio scenes in ASDF in significant manner. The implementation of basic features of the defined format is described. In addition an idea for the sonification of surfaces is described, which uses the definition of trajectories on a surface for the conversation of geometrical data to audio signals. SHORT BIO: Lukas Kaser is a Master's student of Computer Science at the Berlin Institute of Technology. HOST: Sascha Spors Second talk: TITLE: Berührungslose Interaktion am Fraunhofer HHI. SPEAKER: Paul Chojecki HOST: Jörg Müller Last updated 23.11.2011 by Hamed Ketabdar |
| 07.11.2011 | Please notice: there are exceptionally two talks for this date!
Location: Ernst-Reuter-Platz 7, 20th Floor, Auditurium 1
1st talk:
TITLE: A 99$ eye tracker SPEAKER: Youri Marko, EPFL BRIEF DESCRIPTION: Eye tracking is a very exploited technology used in many different research contexts to probe perceptual and cognitive processes. However current gaze trackers are expensive and have strong limitations. Most commercial eye trackers are designed to observe the gaze patterns of users interacting with fixed digital monitors. In addition, the accuracy of these devises decrease rapidly if the subject move during the experiment. As we were interested in analyzing gaze pattern of subjects interacting in a NATURAL way with PAPER documents, solutions that could potentially relax these limitations have been investigated. After testingdifferent remote setups, we built a wearable eye tracking device and tested it in a reading experiment on paper documents. The prototype is constructed with low cost off-the-shelf components. It integrates two cameras and an infrared LED in a pair of glasses. We used a hacked Sony Playstation camera to track the pupil of the subject and a commercial board camera to look at the scene and track the paper document. The results obtained with this first prototype are very encouraging and show that we perhaps don't need to spend several thousands of euros to buy an expensive commercial device in order to conduct eye tracking studies. We really hope that this device will inspire people coming from different fields and allow them to start working on new eye tracking projects.
BIO:
Youri Marko is a research assistant in the/ Peripheral Systems Laboratory/ at /EPFL/ (/Ecole Polytechnique Fédérale de Lausanne, Switzerland/) where he works on printing techniques and color prediction models in order to develop new security solutions for printed documents. He obtained his Master of Science MSc in Microengineering in February 2011 at /EPFL/. During his Master Thesis ("/Design of an Eye-Tracking System for Analyzing Reading Behaviours/"), he developed a 99$ eye tracker. Youri is particularly interested in human-computer interfaces, computer vision, design and image processing. He is currently looking for a challenging position where he could work on interdisciplinary projects."
2nd talk:
HOST: Robert Schleicher TITLE: Multimodal interaction research at 'Tampere Unit for Computer-Human Interaction' (TAUCHI) SPEAKER: Markku Turunen BRIEF DESCRIPTION: I will present an overvirew of multimodal interaction research activities at the Tampere Unit for Computer-Human Interaction (TAUCHI). University of Tampere is one of the largest universities in Finland, with about 14.500 students. It is a full-scale university with nine schools. As a sizeable part of the School of Information Sciences, Tampere Unit for Human-Computer Interaction (TAUCHI) consists of 45 staff members. It is the largest research unit in its area in Finland and one of the largest in the world. The unit it has been identified as the "as the strongest Finnish group in HCI" in the report of Academy of the Finland. TAUCHI is further divided into four research groups, Multimodal Interaction, Emotions, Sociality, and Computing, Speech-based and Pervasive Interaction Group, and Visual Interaction research Group. I will introduce the research activities of these group, and provide some detailed examples from relevant case studies of the Speech-based and Pervasive Interaction Group. HOST: Klaus-Peter Engelbrecht Last updated 19.10 by Hamed Ketabdar |
| 24.10.2011 6 pm -8 pm | Please notice exceptional time! Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 2 TITLE: A Comparison of Eyes-Free Touch Remote Controls for Big Screen Interaction SPEAKER: Jonas Willaredt HOST: Michael Nischt Last updated 12.10.11 by Hamed Ketabdar |
| 10.10.2011 Canceled! | Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 1 TITLE: Multimodal interaction research at 'Tampere Unit for Computer-Human Interaction' (TAUCHI) SPEAKER: Markku Turunen BRIEF DESCRIPTION: I will present an overvirew of multimodal interaction research activities at the Tampere Unit for Computer-Human Interaction (TAUCHI). University of Tampere is one of the largest universities in Finland, with about 14.500 students. It is a full-scale university with nine schools. As a sizeable part of the School of Information Sciences, Tampere Unit for Human-Computer Interaction (TAUCHI) consists of 45 staff members. It is the largest research unit in its area in Finland and one of the largest in the world. The unit it has been identified as the "as the strongest Finnish group in HCI" in the report of Academy of the Finland. TAUCHI is further divided into four research groups, Multimodal Interaction, Emotions, Sociality, and Computing, Speech-based and Pervasive Interaction Group, and Visual Interaction research Group. I will introduce the research activities of these group, and provide some detailed examples from relevant case studies of the Speech-based and Pervasive Interaction Group. HOST: Klaus-Peter Engelbrecht Last updated 26.09.2011 by Hamed Ketabdar |
| 15.09.2011 11:00-12:00! | Please notice: Exceptional date and time! The talk starts at 11:00 sharp! Please notice: The talk is in English Location: Ernst-Reuter-Platz 7, 20th Floor, Auditurium 1 TITLE: Die Effizienz von multimedialen Präsentationstechniken in der Kommunikation des radiologischen Befundes SPEAKER: Franziska Franz, TU Berlin BRIEF DESCRIPTION: In der Radiologie kommen heutzutage hochentwickelte Verfahren zum Einsatz, durch die das Innerste des menschlichen Körpers präzise und kontrastreich abgebildet wird. Dieses Bildmaterial ist Grundlage für den radiologischen Befund, in dem der Radiologe seine Beobachtungen beschreibt und die Beurteilung dazu abgibt. Der überweisende Mediziner, der Empfänger des radiologischen Befundes, erhält in den meisten Fällen einen rein sprachlich- visuellen Befund in Textform. Seit Jahrzehnten erhalten Überweiser Befunde in derselben Form. Das Ausgangsmaterial für seine Therapieentscheidungen ist dasselbe geblieben, obwohl sich das Ausgangsmaterial für die Beurteilungen der Radiologen drastisch verbessert hat und sich viele neue Modellierungs- und Betrachtungsmöglichkeiten bieten. Diese Arbeit untersucht unter kognitionspsychologischen Aspekten, ob Überweiser von einer multimedialen Präsentation der Befundinhalte, die die radiologischen Bilder einschließt, profitieren. Ziel war es, die Effizienz von multimedialen Gestaltungstechniken in der Kommunikation des radiologischen Befundes zu prüfen. Dafür wurden zwei Formen multimedialer Befunde (analog und audiovisuell) mit dem herkömmlichen Textbefund verglichen. HOST: Sebstian Möller Last updated 31.08.2011 by Hamed Ketabdar |
| 18.07.2011 | Location: Ernst-Reuter-Platz 7, 20th Floor, Auditurium 2 TITLE: A Short History of Tabletop Research, Technologies, and Products SPEAKER: Christian Müller-Tomfelde BRIEF DESCRIPTION: In his talk Christian will present a brief history of scientific research into interactive tabletops, associated and emerging technologies, and commercial products. As part of the introductory chapter of the upcoming Springer book about tabletops, it summarizes and visualizes a body of scientific work, identifies major advances during the past fifteen years, and thereby draws a picture of the research landscape to date. Furthermore, key innovations during this period are identified and their research impact discussed. The historical information become synthesized into a synoptic landscape including research highlights, enabling technologies, prototypes, and products. On top of this landscape, innovations are pointed out that stimulated and triggered three key transitions in research and technology. These innovations have also played a major role in leveraging ideas from a conceptual level to widespread adoption and use. BIO: Christian Müller-Tomfelde is a researcher at the Information Engineering Laboratory of the CSIRO ICT Centre in Sydney. His expertise lies in the research area of Human Computer Interaction and Virtual and Hybrid Environments focussing on the support of co-located and remote collaboration support. His interests also include novel forms of interaction, multimodal interaction such as sound feedback. After finishing his studies in electrical engineering at the University of Hamburg-Harburg, he worked at the Center for Art and Media Technology (ZKM) in Karlsruhe. In 1997 he joined the GMD-IPSI's division AMBIENTE "Workspaces of the Future" in Darmstadt/Germany. He was involved in the i- LAND project and in the design of the roomware components of the first and second generations. In his dissertation he explored new forms of audio feedback for the collaborative interaction in hybrid, next generation work environments. As a post-doctoral fellow Christian was investigating aspects of Human Computer Interaction in virtual haptic environments. Now Christian is leading national research projects in distributed collaboration and for interactions on large high-resolution displays and in Multi-Display Environments within the CSIRO. Host: Sebastian Möller Last updated 01.07.2011 by Hamed Ketabdar |
| 20.06.2011 (Cancelled) | Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 2 SPEAKER: Hartmut Pfitzinger Host: Benjamin Weiß Last updated 24.02 by Hamed Ketabdar |
| 30.05.2011 | Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 2 TITLE: Towards Optimized Quality of Experience for 3DTV and Mobile Video SPEAKER: Lutz Goldmann BRIEF DESCRIPTION: Quality of Experience in 3DTV: The success of 3D video, as one of the emerging multimedia formats, will largely depend on the improved quality of experience that it provides to viewers when compared to conventional 2D video. Therefore reliable methods for 3D video quality assessment are crucial in order to optimize 3D video systems and services. The goal of this talk is to review recent developments in subjective and objective 3D video quality evaluation, survey the current research activities of our group in this field and, discuss future directions and possible collaborations. Optimized Video Streaming for Mobile Devices: The rapid technical evolution of portable devices and mobile networks has lead to a time where mobile internet is extremely popular. Considering the usage trends, web video is one of the most important services. The large variety of mobile devices and the limited bandwidth of mobile networks, require an optimization of these video services in terms of quality and resources. The goal of this talk is to review recent developments in mobile video streaming, survey the current research activities of our group in this field, and discuss future directions and possible collaborations. BIO: Lutz Goldmann received the Dipl.-Ing. degree in Electrical Engeinnering from the Technische Universität Dresden (TUD) in 2002 and the Dr.-Ing. degree in Electrical Engineering from the Technische Universität Berlin (TUB) in 2009. Between 2003 and 2009 he worked as a research assistant in the field of multimedia content analysis and machine learning at the Communication Systems Lab, Technische Universität Berlin. In 2009, he joined the Multimedia Signal Processing Group, Ecole Polytechnique Federale de Lausanne (EPFL) as a postdoctoral researcher working on quality of experience, 3D video processing and 2D video transcoding. During the last years he was involved in several national and international projects, including the European Networks of Excellence 3DTV, K-Space, and VISNET and the recent COST Action QUALINET. HOST: Alexander Raake Last updated 24.05 by Hamed Ketabdar |
| 16.05.2011 | Location: Ernst-Reuter-Platz 7, 13th floor, Consilium TITLE: Enabling In-Turn Processing in Spoken Dialogue Systems SPEAKER: Timo Baumann BRIEF DESCRIPTION: Current spoken dialogue systems are not yet suitable for natural, conversational dialogue for a number of technical and theoretical reasons. One major research topic in this area is incremental processing, that is, processing of user input while it is still ongoing. Incremental processing allows a system to react more quickly, to give feedback during a user's turn, or even to interrupt the user. Focussing on the input side, I will present methods that we have developed to improve incremental speech recognition and present ongoing research in predicting when a currently ongoing word is expected to end. HOST: Benjamin Weiß Last updated 11.5.11 by Hamed Ketabdar |
| 09.05.2011 | Location: Ernst-Reuter-Platz 7, 13th floor, Consilium TITLE: Gesture-based Representations of Speech: Acquiring and Analysing Resources for Audio-visual Processing SPEAKER: Stephen Wilson BRIEF DESCRIPTION: I’ll talk mainly about my PhD work which developed set of linguistically motivated gestures that describe visible aspects of the speech production process. These gestures form part of an innovative representational model that allows phonemic sound units to be described in terms of an underlying multitiered gestural structure, one that goes beyond the level of description afforded by the traditional unit of visual speech representation, the viseme. BIO: Stephen Wilson is from Ireland and gained his PhD in Computer Science from University College Dublin for work on speech representations and audio-visual speech resources.At T-Labs he will be working on audio-visual asynchrony and the impact this has on perceived quality. Last updated 5.5.11 by Hamed Ketabdar |
| 08.04.2011 14:00-16:00 | Please notice the exceptional date and time! Location: Ernst-Reuter-Platz 7, 17th Floor, Room Sputnik TITLE: Off-center Sound Degradation Based on Geometrical Properties SPEAKER: Nils Peters BRIEF DESCRIPTION: To improve the reproduction quality at off-center listening positions (OCPs) in surround loudspeaker setups, one needs to understand the nature of the perceived artifacts. Based on the geometrical relationships of a listener to the loudspeaker in a surround setup, an OCP can be defined by three attributes: time-of-arrival differences (ToA), sound-pressure-level differences between the signal feeds, and the direction of the arriving wavefronts. Two listening experiments were designed to elicit the perceptual effects of the off-center sound degradation of each of these three attributes in qualitative and quantitative terms. The five most often qualitatively described artifacts between the tested listening positions are related to the position of sound sources; the distance and depth of sound sources; reverberation and envelopment; spread and width of sound sources; and sound coloration. The quantitative study revealed that off-center sound degradation is primarily caused by the level differences of the loudspeaker feeds. The time-of-flight differences have a stronger perceptual effect on percussive sound material than on sustained sound material. In two out of three musical excerpts, off-center sound degradation was primarily correlated with artifacts related to the reproduction quality of reverb and envelopment. BIO: Nils Peters is a post-doc at the International Computer Science Institute (ICSI) and the Center for New Music and Audio Technologies (CNAMT) at UC Berkeley. He holds an MSc degree in Electrical and Audio Engineering from the University for Music and Dramatic Arts and the University of Technology in Graz, Austria. Recently he defended his PhD thesis at McGill University, Montreal in Music Technology which is entitled "Sweet [re]production: Developing sound spatialization tools for musical applications with emphasis on sweet spot and off-center perception". HOST: Jens Ahrens Last update 28.3, by Hamed Ketabdar |
| 27.01.2011 18:15-19:45! | Please notice the exceptional date and time! Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 3 TITLE: Cognitive Crash Dummies: Where We Are and Where We're Going SPEAKER: Prof. Bonnie John, Carnegie Mellon University Human-Computer Interaction Institute BRIEF DESCRIPTION: Crash dummies in the auto industry save lives by testing the physical safety of automobiles before they are brought to market. "Cognitive crash dummies" save time, money, and potentially even lives, by allowing computer-based system designers to test their design ideas before implementing those ideas in products and processes. In this talk, I will review the uses of cognitive models in system design and the current state of research and practice. I will introduce a tool for easily constructing valid cognitive models of tasks on very early UI design, and present some exciting new research directions that promise to make predictive human performance modeling even more useful. BIO: Dr. Bonnie E. John, a psychologist (PhD, Carnegie Mellon University, 1988) and engineer (BEng, The Cooper Union, 1977; MS, Stanford, 1978), is a full professor in the Human-Computer Interaction Institute at Carnegie Mellon University with has more than 25 years experience in usability analysis and design. She researches both human performance modeling and software engineering and consults regularly in government and industry. Dr. John has been doing research in Human Performance Modeling for 25 years and has published over 100 papers on the topic. She is an ACM CHI Academy member, recognized for her contributions to HCI through her work in cognitive modelling and the implications of usability concerns on the design of software architecture. HOST: Stefan Schaffer Last updated 17.1, by Hamed Ketabdar |
| 26.01.2011 10:00! | Please notice the exceptional date and time! Location: Ernst-Reuter-Platz 7, 20th Floor, Auditorium 1 TITLE: Remediating an urban outdoor space with non-linear psychoacoustics SPEAKER: Don Hill, Liberal Studies, Faculty of Extension, University of Alberta and Behavioural Neuroscience Program, Laurentian University Sudbury Ontario BRIEF DESCRIPTION: Recently published qEEG research measured how human cognition is affected and altered by specific sound derived from a so-called 'sacred place' in Canada's Rocky Mountains. Prominently marked by indigenous peoples 'rock art' -- ancient pictographs, petroglyphs and petroforms -- many of these locations have natural architecture with unusual acoustics; the extraordinary sonic attributes propose contemporary application (inclusive of expanding the frame of reference for LEED building certification). The presenter is currently investigating the psychoacoustic properties of carillon bells (already installed) and how specific tonal frequencies associated with each chime relate to and affect the acoustics of a public gathering space. To remediate the unpleasant sound of a large square in the downtown of a major Canadian city, ongoing research proposes to create tools to optimize the conditions for acoustic standing waves, noise-cancellation effects, as well as introduce specific tones and pulses to create an ongoing “harmony of the square”. HOST: Jens Ahrens Last updated 17.1, by Hamed Ketabdar |
