GVU Demos

in conjunction with

Ubicomp 2001

Sunday, September 30, 2001

6:00 PM - 9:00 PM

As part of the Ubicomp conference held in Atlanta September 30 - October 2, 2001, GVU will host demo sessions in our two main labs, as well as in the Aware Home Laboratory. Following is a listing of demos that will be presented.

Demos in The Aware Home

In collaboration with the Broadband Institute, several GVU faculty are working in a research initiative, known as the Aware Home, where they study how computing technologies can both foster greater independence in the home for elderly people and enhance family connections.

Title	People Involved	Description
AwareHome: Support for Aging in Place	Elizabeth Mynatt Jim Rowan	There can be no denying that America's population is aging. Baby boomers approaching late middle age ask: "How does one care for a population that lives many years longer than any of the preceding generations?" AwareHome: Support for Aging in Place strives to delay taking that first step away from the family home. Through the careful placement of technological support we believe elderly people can remain in their own homes longer.
Aging in Place: Digital Family Portrait	Elizabeth Mynatt Jim Rowan	The digital family portrait provides a qualitative visualization of the general level of activity in the home. Leveraging a familiar household object, our design populates the "frame" of the picture with iconic imagery to depicts the general level of activity in the home for the current day as well as the previous 27 days. If further detail is desired the user can get a more detailed characterization of the day in question by touching the icon that represents the day of interest.
Cook's Collage: Recovering From Interruptions	Elizabeth Mynatt Gregory Abowd Quan Tran Khai Truong	Interruptions distract from the present task such as cooking, and possibly cause irrecoverable memory lapses such as "Did I add the baking powder?" to this flattened batch of brownies? To help alleviate this quandary, the Cook's Collage presents a dynamic synopsis of recent activity within a room, namely the kitchen. This always-on visual display provides serendipitous information to the user who would not notice it while attending to his present task, but will rely on the helpful collage to recover from interruptions.
Encouraging Value-Responsive Design	Elizabeth Mynatt Amy Voida Kevin MacDonald	The use of technology is motivated by the values that undergird the social practices in which it is used. If HCI methods are to impact the usefulness of technology, they, too, must be able to expose and be responsive to the values of the intended users. We are developing a methodology, both data collection and design process, for designing technologies for families in home environments. Our methodology is based on a framework of understanding family values. Our data is collected through families' collaborative creation of specified artifacts (value probes) that allow family members to express situated values. Our research regarding how values are manifested in family life is closely coordinated with research regarding how to enable designers to work constructively and creatively with values-based data. The later component of our research yields a step-by-step design process, tailored to the specific needs of designers working with families and value probe artifacts.
The Family Intercom: Supporting Context-Aware Communications	Gregory Abowd Kris Nagel	There are many technologies enabling direct human-human audio communication, but none of them use the context of the communication partnerseffectively. We are augmenting a domestic environment with implicilty gathered knowledge of the location and activities of its occupants. The Family Intercom provides a test bed to explore how this context supports a variety of lightweight communication opportunities between collocated and remote family members.
RF ID Indoor Positioning	Gregory Abowd Peter Jensen Thomas O'Connell Arpit Agarwal Venkat Ramachandran	We utilize Radio Frequency Identification (RFID) systems in the Georgia Tech Aware Home to sense the location of participating persons to facilitate contextually aware applications. RFID antennas fashioned into floor mats and passive transponders attached to participant's shoes provide the basis for this system. We attempt to use consecutive mat readings to interpret room level location
The Discoverer	Gregory Abowd Agathe Battestini	The Discoverer is a discovery system built on top of the Context ToolKit. In a context-aware environment, sources of contextual information are likely to be modified, added, and removed. The Discoverer allows applications to be constantly aware of these changes by notifying them of important events.
Audio Infrastructure	Gregory Abowd Kris Nagel Andy Wightwick Gary Schwaiger Sourabh Ravindran David Anderson	The Audio Infrastructure provides audio services between rooms within the house and implicitly gathers context information. Digital signal processing for speaker identification both names and locates an individual within a room. The separation of two speakers within the same room into distinct audio streams is another active research topic. We are also investigating a variety of protocols to transport voice between rooms within the home and between homes.
Visual Tracking	Aaron Bobick Jake Auxier	One of the foundations of the perception of those living in the home is determining where they are located in the home. Currently, only room-level location information is available. Although room level location information is useful in many applications, it remains very limiting. More interesting applications and research can be done if better location information can be provided that is able to tell where somebody is located in a particular room. One method of providing this information is through computer vision. An object tracking system is being developed to provide this information through the use of cameras mounted in the ceilings of the Aware Home. These cameras provide overhead views of the home. The object tracking system will be able to provide not only more specific information on where a person is located in the home, but it can also provide orientation of the person being tracked.
Multi-modal Sensing in Aware Environments	Irfan Essa Scott Stillman	Most applications within any form of a smart environment (rooms, offices, homes, etc.) are dependent on reliable who, where, and when, and what information of its inhabitants (users). This information can be inferred from different sensors spread throughout the space. However, isolated sensing technologies provide limited information under the varying, dynamic, and long term scenarios (24/7), that are inherent in applications for intelligent environments. This work provides an infrastructure for leveraging the strengths of different sensors and processes applied to their collective data.
Peek-A-Drawer	Itiro Siio Elizabeth Mynatt James Rowan	Grandparents want to share a part of their living space with their grandchildren who live far away. Peek-A-Drawer provides virtual shared drawers across distant locations. When a user puts something in the upper drawer and closes it, a photograph is taken automatically, and the image appears in the lower drawer at a distant place. Past pictures can be accessed by pressing buttons on the lower drawer. The operation is so simple -using a drawer- that even children can communicate with their grandparents. As the camera only takes objects inside the drawer, privacy is assured.
Meeting Pot	Itiro Siio Noyuri Mima	Community building in coffee break rooms is important in modern offices. Simply the aroma of coffee evokes togetherness. The Meeting Pot attempts to distribute this sense of awareness. When the coffee maker in the kitchen of the Aware Home is turned on, it transmits the aroma to remote locations. In order to encourage informal communication in the Future University of Hakodate (Japan), we have installed a Meeting Pot in a common room at the faculty building. These boxes blow a coffee aroma through the lid when the coffee maker is turned on.
Low Cost Indoor Positioning System	Cliff Randell Henk Muller University of Bristol, UK	This system simulates the progress of a fire in a single family dwelling and responds to actions made by the user to rescue occupants and put the fire out. The user of the VE is a Fire Company Officer being trained or evaluated for his/her skills at commanding a fire crew. In the VE, the user will speak commands that are translated by an operator into a predetermined animation sequence in the virtual environment. As the fire company officer issues commands, the virtual fire crew will go through animations reflecting these commands, and the fire burns in response to virtual crew actions.

Demos in the College of Computing Building

Title	People Involved	Description
Temporal Integration of Multiple Silhouette-based Body-part Hypotheses	Aaron Bobick Vivek Kwatra Amos Johnson	We present a technique for integrating appearance-based body-part labeling. It works as a probabilistic tracking framework extended to handle multiple competing state estimators.
The Kimura System: An Augmented Office Environment	Elizabeth Mynatt Blair MacIntyre Gregory Corso Stephen Voida	The Augmented Office seeks to explore and evaluate the addition of visual peripheral displays to human-computer interfaces within an office environment. Our research is focused on augmenting a traditional desktop display with semantically meaningful, task-based wall displays, as well as exploring the psychological principles underlying perception of change in a peripheral display.
Ambush	Elizabeth Mynatt Joe Tullio	A probabilistic calendar portrays the likelihood of a person attending an event based on the content of their calendar and their past attendance patterns. Using Bayesian networks, we model attendance patterns based on a number of factors including the existence of conflicting events, past behavior, and the relationships between repeating, unique, incomplete and recent calendar entries. This research investigates the applicability of Bayesian nets to predicting attendance, how Bayesian models can adapt to naturally changing behaviors such as declining interest in a recurring event, and how sporadic training and automatically gathered information can aid reinforcement learning.
Fostering Group Awareness Through Comparative Representations	Elizabeth Mynatt Elaine Huang	The goal of this research is to build a set of tools to support awareness of others in an office/lab environment, within a set of coworkers or between groups of coworkers. The aim is to provide abstract representations of individuals' status on projects, papers, or websites on a public display, allowing the individuals to assess their own progress in relation to their coworkers' and gain an awareness of the group's activities in general, but also to provide a measure of privacy by not giving individuals access to specifics on other individuals' activities.
Ubi-Finger: Gesture Input Device for Mobile Use	Koji Tsukada Michiaki Yasumura Graduate School of Media and Governance, Keio University, Japan	This paper proposes a new interface in mobile environment called "Ubi-Finger" that realizes sensuous operations for PDA and information appliances by human gesture. Since gesture-input interfaces enables sensuous operations for users, there have been many researches about them especially for Virtual Reality. But almost those existing systems are very expensive and large, and not considered to be used in mobile environment. Ubi-Finger is a gesture-input device, which is simple, compact, and optimized for mobile use. We developed a prototype that enables to control real-world devices with natural gestures.
TrekTrack: A Round Wristwatch Interface for SMS Authoring	Anders Kirkeby Rasmus Zacho Jock Mackinlay Polle Zellweger Center for Pervasive Computing, University of Aarhus, Denmark	The user interface for text messaging via SMS has changed little since the technology was introduced on cell phones. In the part of the TrekTrack project presented here we suggest a future alternative user interface for SMS messages based on a round wristwatch device.
Making Everyday Life Easier Using Dense Sensor Networks	W. Steven Conner Intel Architecture Labs, Intel Corporation Lakshman Krishnamurthy Intel Architecture Labs Roy Want Intel Research	Advances in hardware are enabling the creation of small, inexpensive devices and sensors. Hundreds or thousands of these devices can be connected using low-power multi-hop wireless networks. These networks foster a new class of ubiquitous computing applications called proactive computing. In proactive applications, computing occurs in the background without requiring human interaction; humans participate to access information or to modify control policies. We demonstrate the application of large wireless networks of sensors to solve everyday problems in the workplace, e.g., an application that allows people in the workplace to easily find empty conference rooms (e.g., for impromptu meetings) or to find empty parking spaces in a large parking lot.
Multibrowsing: Moving Web Content Across Multiple Displays	Brad Johanson Shankar Ponnekanti Caesar Sengupta Armando Fox Stanford University	Multibrowsing is a framework that extends the information browsing metaphor of the Web across multiple displays. It does so by providing the machinery for coordinating control among a collection of Web browsers running on separate displays in a ubiquitous computing environment. The dis- plays may be "public" (e.g. wall-sized xed screens) or "private" (e.g. the screens of individuals' laptops or handhelds). The resulting system extends browser functionality for existing content by allowing users to move existing pages or linked information among multiple displays, and also enables the creation of new content targeted specially for multi- display environments.
Lessons Learned in Creating Real-World Interfaces	David Wrighton Dillon Bussert Scott McCrickard Virginia Tech	Informational displays have been developed that use lighting, air flow, and physical objects external to the computer screen, but typically lacking are simple and straightforward steps for creating these displays. This paper describes our experiences in creating a real-world interface using X10 devices and common household appliances, and outlines our framework for creating real-world interfaces quickly, easily, and inexpensively.
Push-Pin Computing: an Integrated Platform for Ubiquitous Computing	Joshua Lifton Joseph Paradiso MIT Media Lab	Pushpin Computing is the first attempt at a computing architecture closely following the ideas set forth in the Paintable Computing project. Namely, it is an architecture consisting of many independent, locally communicating computing nodes physically distributed at scales considered dense in comparison with normal human scales. The key to this model is the idea that simple local interaction among computing nodes can result in complex algorithmic structures at the level of the system as a whole. Strong analogies exist between Pushpin Computing and biological systems such as ant colonies, thermodynamic systems such as ideal gases, and abstract systems such as the game Go.
Video-based Rendering	Irfan Essa Gabriel Brostow Antonio Haro Lionel Reveret Arno Schoedl	We will show various projects on video-based rendering. These includes our work on (a) generating novel video sequence from shorter bounded videos (video textures); (b) synthetic rendering of motion-blur as a video post-process; (c) modeling and rendering of facial skin; and (d) modeling of speaking faces. Time permitting, we will also show our other work on tracking, and may also show videos of some of our Digital Video Special Effects classes.
The Real-world Wide Web: Integrating the WWW into the Real World	Blair MacIntyre Rob Kooper	This project is an exploration of how one would visualize and interact with the WWW if we used it to create a context- and location-aware mobile information space that was projected into 3D around the user as they moved through the world.
Alices Adventures in New Media	Blair MacIntyre Jay Bolter Maribeth Gandy Emmanuel Moreno Joaquin Lobo Madruga	Alice�s Adventures in New Media is an Augmented Reality (AR) experience developed in collaboration between the School of Literature, Communication and Culture, and the College of Computing. The experience is based on A Mad Tea Party, a chapter from Lewis Carroll's book Alice's Adventures in Wonderland. The user assumes the role of Alice and sits at the tea party with three interactive characters: the Mad Hatter, Dormouse, and March Hare. As a developing medium, AR has yet to establish itself as a narrative form. By comparing the unique characteristics of AR with established media such as film and interactive CD-ROMs, the project explores the development of AR as a storytelling medium. Innovations include the refashioning of conventions used in film and interactive CD-ROMs for the development of an AR narrative, the use of simple procedural characters to create an immersive interactive experience, and the exploration of the Macromedia Director environment as an AR production tool. We are also using this demo as a test of a wide area vision tracking infrastructure we have recently begun working on.
An Audio Museum Tour of Sweet Auburn	Blair MacIntyre Jay Bolter Lauren Keating Frances Hamilton Russell Morris Kiana Tennyson	This project is an Audio Augmented Reality (AR) experience developed in collaboration between the School of Literature, Communication and Culture, and the College of Computing. The goal of this project is to explore how spatialized audio can be used to enhance a museum experience by creating an auditory space that envelops the user as they move about the museum, rather than simply playing linear audio clips for each piece or exhibit.
AquaMOOSE 3D	Amy Bruckman Jason Elliott	AquaMOOSE 3D is a graphical multi-user game construction kit that provides a compelling context for high school students to learn mathematics. The current version includes the Ring Game, where students travel through an obstacle course of rings in the virtual space by specifying their movements with parametric equations.

Demos in the Centennial Research Building, 2nd Floor

Title	People Involved	Description
Space Cadet - New Techniques for Window Management	John Stasko Dugald R. Hutchings	We are exploring algorithms for efficient management of windows in GUIs with a focus on developing new movement/repositioning operations.
What's Happening - Promoting Community Awareness	John Stasko Alex Zhao	Keeping up with the current affairs in a community and knowing community members beyond their names is difficult in large, somewhat virtual organizations. This demo will present two lightweight opportunistic information displays that help promote awareness and communication without causing distraction.
Organize This! Exploring Personal Information Management	John Stasko Rod Peters	This demo will present the results of a qualitative research study aimed at better understanding people's habits in personal information management. The results of this study are being used to inform the design of an information visualization system for helping people in this activity.
Webwatch: Exploring Visualization Techniques for Viewing and Sharing Bookmarks	John Stasko Cathy Eichholz Polk	The Webwatch research project investigates information visualization techniques for displaying bookmarks of individuals and groups of people. The visualization aims to facilitate the sharing and discovery of interesting web pages among peers.
InfoCanvas - Information Art	John Stasko Todd Miller	This demo shows a system for helping people keep aware of information of peripheral interest. Users "paint" a personalized representation of information of interest (traffic, weather, etc.) and we display the appropriate mapping of that representation on ambient LCD displays.

Demos in the Centennial Research Building, 3rd Floor

Title	People Involved	Description
Automated Meeting Capture and Access in TeamSpace	Gregory Abowd Heather Richter	TeamSpace is a collaborative environment for distributed teams supporting the capture and access of virtual meetings. This is a demo of the system described in the Ubicomp 2001 paper "Integrating Meeting Capture and Access within a Collaborative Environment."
Multimodal Interfaces	Bill Ribarsky Thad Starner Chris Shaw David Krum	We demonstrate our new speech and hand gesture multimodal interface. The interface integrates speech recognition and vision-based hand gesture recognition. We apply the interface to a rich navigation application in a whole earth environment. The interface is flexible and can be used on a variety of platforms. We demonstrate it in our mobile visualization and virtual workbench environments.
GVU Weather Central	Bill Ribarsky Nick Faust Chris Shaw	We demonstrate our integrated environment that obtains real-time, 3D Doppler radar feeds and satellite weather images and displays them immediately over high resolution terrain. A variety of display interfaces and visualization techniques are used to support rapid decision-making. The goal is a ubiquitous data storage, query, and decision-making environment where a variety of stationary and mobile users can get severe storm weather information for the location they want whenever they want.
The Virtual Airplane	Larry Hodges Chris Shaw Jeff Stewart	Fear of flying is a serious problem that affects millions of individuals. Exposure therapy for fear of flying is an effective therapy technique. However, exposure therapy is also expensive and logistically difficult to arrange. To solve these problems, we have developed a virtual airplane for use in fear of flying exposure therapy.
The Meditation Chamber	Larry Hodges Diane Gromala Chris Shaw Jay Bolter A. Fleming Seay Mirtha Ferrer Eli Wendkos Sue Rinker Chris Aquino Chris Campbell Colin Henderson Rob Hill Dawn Pendergast	The Meditation Chamber is a Virtual Environment experience that is intended to promote relaxation through the use of biofeedback visualization, progressive muscle relaxation, and controlled breathing. Users wear a head-mounted display and are taken through a series of relaxation exercises, coached by audio and visual imagery.
Fire Training Environment	Chris Shaw Larry Hodges Mahesh Balakrishnan Tazama St. Julien Robert Crawford	This system simulates the progress of a fire in a single family dwelling and responds to actions made by the user to rescue occupants and put the fire out. The user of the VE is a Fire Company Officer being trained or evaluated for his/her skills at commanding a fire crew. In the VE, the user will speak commands that are translated by an operator into a predetermined animation sequence in the virtual environment. As the fire company officer issues commands, the virtual fire crew will go through animations reflecting these commands, and the fire burns in response to virtual crew actions.
The Gesture Pendant	Thad Starner Daniel Ashbrook Maribeth Gandy Helene Brashere Marty McGuire	We present a wearable device for control of home automation systems via hand gestures. This solution has many advantages over traditional home automation interfaces in that it can be used by those with loss of vision, motor skills and mobility. By combining other sources of context with the pendant, we can reduce the number and complexity of gestures while maintaining functionality. As users input gestures, the system can also analyze their movements for pathological tremors. This information can then be used for medical diagnosis, therapy and emergency services. Currently, the Gesture Pendant can recognize control gestures with an accuracy of 95 percent and user defined gestures with an accuracy of 97 percent. It can detect tremors above 2HZ within +-0.1 Hz.
Billiards Assistant	Thad Starner Tracy Westyn Amy Hurst	The Billiards Assistant analyzes the current board and players' shot history to recommend which shot to make next.
Capture Vest	Thad Starner Kent Lyons	In order to study human-computer interaction on a wearable computer, the physical context of the situation must be captured. The Capture Vest and VizWear software allows replaying of both the user interface and the user's context for user studies.
Guardian Agents	Thad Starner Ben Wong	Wearable computers are often promoted with speech recognition interfaces that allow the user to select options from a menu. However, our lives are surrounded by the spoken word. What speech interfaces are viable outside that of traditional direct manipulation? This project attempts to couple everyday conversation with visual memory aides such as the Remembrance Agent and schedulers to provide just-in-time information support for the user.
Detecting Social Engagement with Wearable Computers	Thad Starner Brad Singletary	Description TBA
Mobile Sign Language Recognition	Thad Starner Helene Brashere	We are extending previous research on a portable American Sign Language (ASL) translation system that can be used in a real world environment. The system will translate the user's sign into spoken words. It is composed of a wearable computer and a video camera mounted in a hat. The video camera feeds images of the users hands to the wearable computer where it is processed and interpreted.
Animal Behavior Modeling	Ron Melby	Description TBA

Research | Education | People | Labs | News & Events

GVU Demos

in conjunction with

Ubicomp 2001

Sunday, September 30, 2001 6:00 PM - 9:00 PM

Demos in The Aware Home

Title

People Involved

Description

Demos in the College of Computing Building

Title

People Involved

Description

Demos in the Centennial Research Building, 2nd Floor

Title

People Involved

Description

Demos in the Centennial Research Building, 3rd Floor

Title

People Involved

Description

Sunday, September 30, 2001

6:00 PM - 9:00 PM