Microsoft Research announced the five recipients of the ConferenceXP Research Platform awards, totaling $250,000 in funding. The objective of the ConferenceXP Research Platform award is to help further research and teaching in the areas of real-time collaboration, wireless-enabled classrooms, and distributed learning. The ConferenceXP research platform extensions and ConferenceXP applications designed at the following universities will further enhance learning and research processes and will advance the state of the art within their domains.
Enhancing Reliability by Supporting Path-Diversity Overlay Retransmission
University of Missouri at Columbia
Summary This project investigates a path-diversity overlay retransmission architecture and mechanism to achieve more effective and reliable packet delivery in challenging network environment This project investigates the retransmission architectures and mechanisms to achieve more effective and reliable packet delivery in challenging network environments. The more traditional IETF RTP retransmission protocol will be implemented and its performance will be evaluated. More importantly, a novel overlay retransmission architecture that exploits path diversity will be investigated to improve or complement the traditional approach.
In the proposed path diversity overlay retransmission architecture, a very simple overlay retransmission network is built among the participants. Each receiver identifies a couple of good ï¿½retransmission nodes,ï¿½ which will retransmit/forward the packet they receive to the requesting receiver who experiences packet loss through a different path, resulting in improved reliability and load balancing. The main thrust of the proposed overlay retransmission architecture is its ability to exploit path diversity to reduce the retransmission delay and to avoid the potential need for the original sender to retransmit a lost packet multiple times due to a severely congested path. The project is expected to significantly enhance the capability of the ConferenceXP platform beyond its current limitations.
Visual Information Manager
University of Illinois at Urbana-Champaign
Summary We will design and implement a Visual Information Manager as an extension to ConferenceXP Platform that manages the screen display resource and coordinates the user interface presentation between video, presentation, and other visual channel Multiple visual information channels are presented on the ConferenceXP platform, such as video streams from live cameras, PowerPoint presentation, whiteboard, and Web cobrowsing. Therefore, the screen space for presenting the interface window of these channels becomes a critical resource that needs to be well managed. Todays ConferenceXP employs a simple tiled window mechanism to spread the video windows on the screen, which does not scale well in the case where multiple camera streams from multiple sites and other applications, such as PowerPoint presentations, are presented.
In this project, we propose to design and implement a Visual Information Manager as an extension to ConferenceXP Platform that manages the screen display resource and coordinates the user interface presentation between video, presentation, and other visual channels. Specifically, we focus on the following three challenges. First, we will propose a model of the available display resource at each conference site, such as the resolution and number of display devices. Second, we will model each visual information channel in terms of its display resource requirement, and the user interest level in each channel based on the type of information it provides over time. Third, we will propose a display resource manager module that allocates display resource in real time so that visual channels that carry more interesting information will be selected and their interface windows are then composed into a nice screen layout.
Enhancing Cooperative Learning with ConferenceXP-powered I-MINDS
University of Nebraska
Summary My project is to empower a computer-supported cooperative learning system called I-MINDS using the ConferenceXP platform.
Abstract The long-range goal of our research is to improve teaching and learning through technology, where agents support cooperative learning among students and support teachers in managing real-time classroom activities. Other goals include bringing the same quality of interaction and team building to distance education for both synchronous and asynchronous learning and addressing the challenge in teaching and student performing well in introductory CS courses (for example, CS1 and CS2). Towards accomplishing the above goals, we have developed and prototyped a new technology: Intelligent Multiagent Infrastructure for Distributed Systems in Education (I-MINDS). Though I-MINDS has been prototyped and pilot tested successfully, it is faced with several inhibiting problems in terms of the enabling technologies for it to be deployed widely. These problems include the lack of readily stable and scalable conferencing capabilities, lack of security and firewall handling, inefficient digital archival of lectures and notes, and insufficient user modeling. Therefore, we propose to:
Windows Media Transcoding Using ConferenceXP Archive Service Data
University of Washington
Summary Develop tools to transcode ConferenceXP Archive Service data including audio, video, and presentation into Windows Media forma Conferences and classes archived by the ConferenceXP Archive Service will gain value if they can be viewed in a scalable fashion by clients with lower network bandwidth capabilities, by clients behind firewalls, and by disconnected clients. The preferred archive format for these scenarios is Windows Media. This project will address the transcoding of data stored by the ConferenceXP Archive Service into Windows Media format.
The output of the transcoding process will include not only audio and video, but also presentation slides, navigation, and ink, and will interoperate well with existing applications, such as Classroom Presenter and ConferenceXP WebViewer.
Advanced Classrooms Exploiting Tiled Displays and Student Computers
University of California at Santa Cruz
Abstract We are experimenting with use of multiple (large and multi-mega pixel) tiled displays, allowing all lecture visuals to remain in view (and downloadable) during the lecture, and protocols for the instructor to enable and manage presentation of selected student responses and results. In addition, we will develop high-resolution and bandwidth-efficient support for high-resolution document cameras as an input to ConferenceXP.
Summary In spite of a vast government sponsored network of health/medical facilities, many villages in India, remain deprived of the benefits of modern medicine. There are many reasons for this which, are well documented elsewhere. The area we wish to address is ignorance and superstition about the cause of disease, patient care, and related issues. This exists primarily due to lack of information. Illiteracy has, in the past, contributed to the problem. However, with access to digital technologies it is now possible for interactive multimedia content to be used to cross this barrier.
We have developed a package under a Media Lab Asia initiative. This consists of:
ConferenceXP, in its functionality, appears suitable for the deployment of the modules developed by us and our application will extend it to domains beyond the classroom/academic environs and demonstrate its applicability in a different linguistic, instructional, and cultural context.
Summary We propose developing a mixed-reality tabletop (MRT) for immersive mentoring on top of the ConferenceXP research platform. The MRT concept goes beyond the traditional white-boarding concept and allows for a mentor and remote mentees to simultaneously demonstrate operations on real or virtual objects on the same tabletop. Moreover, such a mentoring session can take place without prior scripting and complex object modeling. The MRT concept is highly versatile and can be applied to immersive mentoring in a wide spectrum of learning environments, ranging from K-12, college, to technical training.
The MRT system requires significantly less infrastructure, compared with a fully immersive virtual reality system. Different from a traditional video conferencing system, MRT creates one common plane for immersive demonstration and perception, so that users do not have to shift their vision and attention back and forth between the computer screen and the physical objects. We propose a simple MRT hardware setup, as well as a MRT software toolkit based on the ConferenceXP research platform. We will perform MRT performance evaluation and optimization, as well as user study in collaboration with our K-12 Outreach Program partners.
Summary These days its easy to try mathematical exercises on a computer, as there are many available software packages for that purpose. But most of them are designed to work out of the classroom. Our objective is to develop a mathematical whiteboard to teach mathematics and calculus at different locations, integrating teachers and students using the ConferenceXP platform, mainly for teaching in engineering and applied sciences undergraduate courses. This software will be designed to draw 2-D and 3-D images of mathematical functions, and mathematical and engineering tools (like root finding, Fourier transform, among others) over a ConferenceXP platform, and allow a teacher to reach students at a remote location over the Internet in real time.
Summary Established as a research project at the University of Illinois in the spring of 2002, e-Fuzion has proven to be a valuable and effective pedagogical set of tools. It provides the capacity to both mentor and assess students individually, both during and outside of class. In the summer of 2002, a study we conducted revealed that e-Fuzionï¿½s in-class tools increased studentï¿½s final grade by more than six points. This proposal is focused on allowing those tools to extend beyond the scope of the classroom, and we believe the ConferenceXP platform is the best way to accomplish this task.
Summary The goal of the ReMarkable Texts project is to provide a compelling student experience. This component will be the primary document organizer and viewer for classroom slides and video, handouts, student notes, and academic articles. In addition, ReMarkable Texts will support a range of both private and synchronously or asynchronously shared annotation capabilities, including Ink mark-ups, text highlighting, document hyperlinks and overlays, bookmarks, and context-specific displays.
Summary Experiments and lab work are important components in Engineering and Computer Science courses. Students learn practical concepts and get hands-on experience. Usually students do computer or lab work at a specific time. During these times one or more Teaching Assistants (TAs) are working onsite to help students with their work. However, an onsite TA might not be available (as in the case of the new UC Merced campus) or help may be needed outside of normal lab hours. In this proposal, our goal is to first develop a basic Remote TA Support System leveraging the ConferenceXP platform. Using high-quality and low-latency live video provided by ConferenceXP, a remote TA shares the same visual context with the students, thereby improving interaction and help. In addition to the live audio and video, we plan to extend the Remote TA system by introducing a number of additional features. These include screen sharing, a group monitoring facility, session meta-information, virtual zoom, and stored video delivery.
Summary The promise of teleconferencing systems is to support distributed meetings, collaborative work, seminars, lectures, tutorials, and training by using the facilities provided by the next-generation Internet. Teleconferencing systems are a combination of a number of different technologies: audio and video compression, capture and display, networking and real-time systems. These technologies have all progressed much in recent years, leading to widespread deployment of conferencing. There are, however, two potential problems that may hinder the widespread deployment of teleconferencing systems: divergence in support for codecs and media formats, and the lack of widespread congestion control. This has the potential to both disrupt the conferencing experience and to cause instability and congestion in the Internet. Neither of these are desirable outcomes, and this projects aims to address these issues.
Summary The project investigates next-generation collaborative teaching and learning using ConferenceXP. It builds on the concepts established by the Learning Experience Project and goes one step further into a far more enhanced interactive and multi-media environment. The proposed effort links experts from education, graphics, visualization, and networking, and will conduct research on course delivery at UC Davis linking to UC Santa Cruz. While the tool developed in this project is applicable to all courses, we chose a chemistry laboratory course as an example. CITRIS investigators and the leadership at UC Davis have shown very strong interest in this project, and we expect that CITRIS campuses (Davis, Berkeley, Santa Cruz, and Merced) will pursue collaborative teaching and student conferences over the developed tool for a state-wide demonstration.
Summary The Teacher Experience Project continues work that was initiated in the Learning Sciences and Technology group at Microsoft Research in 2001-2002. This project will promote the use of technology in the classroom to enhance presentation and enable a broader range of in-class activities, with the goal of making instruction more effective by giving the instructor more flexibility in the classroom.
TeleEducation/TeleCollaboration and Streaming Multimedia
Patrick Mantey, J. J. Garcia-Luna, Hai Tao, Sireesh Potireddy, Dan Kong, Jeremy Richards, Liz Watt, Jon Webb
University of California Santa Cruz
Summary We proposed to develop and improve tools and techniques for distance learning by using the Internet in engineering education, supporting remote classrooms synchronously linked by the Internet2, and providing synchronous support of interactions between faculty and students in the classroom and as ï¿½office hoursï¿½. We also proposed to develop protocols for managing multi-party interaction and algorithms to improve the support for person-to-person interaction via the Internet, synthesizing eye contact for improved communication.
In the first year we have made significant progress, even with a slow start as we assembled our research team, obtained lab space, and acquired and installed equipment. The research prototype software from Microsoft Research, called ConferenceXP, is the basic software for this research, now enhanced by our own additions for protocol management. In addition, we have experimented with other complementary software for interactive use in teaching, such as the Classroom Presenter software from the University of Washington, and also with a variety of cameras and input devices. We now have a working experimental facility for our research, with a lab in the Baskin Engineering building (BE146) and use of the multimedia classroom (BE156), plus an installation at the University of California Santa Cruz (UCSC) Silicon Valley Center at NASA Ames. In addition we have experimented with these tools in collaborations with colleagues at UC Berkeley and UC Davis. Two UCSC engineering courses have provided us with a test environment for live streaming, archiving and multimedia database, and this is expected to grow.
Summary The proliferation of wireless 802.11 Local Area Networks (LANs) into university campuses makes these LANs viable communication platforms for a wide range of multimedia learning applications. Hence, the Microsoft ConferenceXP research platform can benefit significantly from solutions that will enable a successful deployment of rich ConferenceXP multimedia applications over emerging wireless LANs. Congestion control, rate control, and low-delay error-resilient networking are examples of key technologies needed for a successful deployment of emerging wireless multimedia applications. Building upon our extensive research efforts in wireless multimedia and congestion/rate control, the objective of this project is two fold: