Microsoft Research Awards Nearly $1 Million to Academic Researchers

REDMOND, Wash., Feb. 21, 2006 – Frank Dellaert hopes someday to make virtual time travel a reality by being able to create 3-D models with time sliders using digital images. Frank Piessens wants to develop not just a few study courses on Trustworthy Computing, but an entire track of graduate-level curriculum devoted to secure software development. What do the two men have in common? As winners of Microsoft’s Request for Proposal (RFP) process, they each will receive up to US$50,000 to help make their visions real.

Dellaert and Piessens are two of 23 academic researchers whom Microsoft named this week as winners of the Microsoft Virtual Earth and Trustworthy Computing RFP programs. The company will award up to $50,000 to each recipient, for a total of more than $1 million. The Microsoft RFP program is the largest in the industry, having funded more than 125 research projects and awarded academic institutions worldwide over $4 million to encourage innovative research and curriculum development in many areas of computing, including social computing and gaming.

“Very few companies have the vision or the resources to support long-term research like this, so programs like Microsoft’s Virtual Earth RFP program are critical in enabling academia and the IT industry to work together and think outside the box,” says Dellaert, an assistant professor at the Georgia Institute of Technology in Atlanta.

“Improving software quality and security plays a significant role in increasing the trust people have in information technology,” says Piessens, a computer science professor at the Katholieke Universiteit Leuven in Leuven, Belgium. “Microsoft’s Trustworthy Computing RFP program is playing a significant role in ensuring computer science graduates have the knowledge and skills they need to build high-quality and secure software systems.”

Advancing the State of the Art of Digital Photography

The Virtual Earth RFP encourages university research in areas relevant to digital photography, such as spatial and spatio-temporal databases, computer vision and visualization. It is a collaboration among university research labs, the Virtual Earth product group and Microsoft Research that encompasses three Microsoft efforts: MapPoint, a set of integrated products and services that helps people harness mapping and location technologies; MSN Virtual Earth, an online virtual globe mapping platform that provides access to satellite imagery and aerial photography of the earth; and the World-Wide Media eXxchange, an experimental research project that features a centralized index of digital photos, tagged by the geographic location where they were shot. The eight winners of the program will focus their research on one or more of these components.

Dellaert will work with students who specialize in the areas of robotics and computer vision to execute his “4-D Cities” research project. Their goal will be to automate the construction of a spatio-temporal model – a model that incorporates space and time dimensions – showing the evolution of a particular city using a large collection of images. Dellaert believes that time-varying 3-D models like the one his team will attempt to create have the potential to pull together and index large collections of images that relate to one place or artifact, including its appearance, evolution and even surrounding events at a given point in time.

Such a technology would enable people to surf back in time using a virtual globe-mapping platform like Microsoft Virtual Earth to see how a city like New York, for example, has evolved over time, how a particular building has evolved over time, or what environment their ancestors grew up in and how it has changed over time. A time slider also holds the potential to allow Web-site visitors to click on a model of a building that no longer exists and do things like retrieve all the newspaper articles that have been written about it in the past.

“Winning this award will allow our team of researchers to dig deeper and really explore new ways of interacting with large databases of images,” says Dellaert. “We look forward to working closely with Microsoft Research on this technology.”

Dellaert’s project, like several others in the Virtual Earth RFP winners’ circle, is sufficiently forward looking that it may be some time before the technology could actually be integrated into Virtual Earth products and services. However, the goal in choosing Virtual Earth RFP winners, according to Kentaro Toyama, assistant managing director of Microsoft Research India, was to end up with a mix of research directions that in some cases may provide solutions to today’s challenges, and in other cases would explore new areas not yet imagined.

“It would be great to see the research results from this year’s RFP proposals eventually integrated into Virtual Earth and Virtual Earth products and services, which impact hundreds of millions of users,” Toyama says. “But it’s difficult to predict which ones will have an impact in the short term. What we do know is that the RFP winners will contribute to the ongoing dialogue in the research community that pushes the state of the art forward.”

Other winning projects, such as Dieter Fox’s “personal location maps” proposal, have the potential to impact Virtual Earth in the short term.

Fox, an associate professor at the University of Washington in Seattle and director of the Robotics and State Estimation Lab, wants to develop techniques for generating personalized maps from traces of location data. By being able to extract information about a person’s location from different types of sensors, such as from GPS devices, WiFi technology and GSM signal strength, it would be possible to generate personalized maps that contain automatically learned, user-specific information, such as where a user works and lives, which routes he or she usually takes to get from one location to another and what type of route he or she is likely to prefer when traveling to a new location, based on past history.

Such maps could provide smart navigation guidance – for instance, the system could anticipate the route a user will drive to work and automatically suggest alternate routes if there’s a traffic delay on the usual route, or it could provide a suggested route to a new location based on the types of roads the user tends to prefer and what segments of the route he or she is likely to be familiar with.

“This type of research project enables the IT industry to provide better solutions in a relatively short period of time,” says Fox. “Designing systems that can automatically help users navigate based on personal preference is just the start of a whole flurry of location-aware capabilities that can provide more focused services for people.”

Ensuring Trustworthy Computing Curricula Is Available for Next Generation Scientists

It’s no secret that the world needs more highly skilled professionals who can design and deliver high-quality, secure and reliable software. A recent President’s Information Technology Advisory (PITAC) report “Cyber Security: A Crisis of Prioritization,” pointed out major deficiencies in the way software is created, and made special note of the lack of qualified academic experts in the field relative to the scope of the problem. This shortage of qualified professionals has a direct impact on the quality and quantity of curricula material available to teach the next generation of scientists and policy experts.

The Department of Computer Science at Katholieke Universiteit Leuven in Belgium was one of the first universities to integrate secure software development into its computer science curriculum. It introduced a dedicated secure software development course in 2001 and now, with the help of Frank Piessens, is developing an entire secure software track as part of its computer science graduate program.

The immediate goal, according to Piessens, is to create a number of reusable teaching modules on recent research results and technologies that have already had an impact on actual software security or can be expected to have an impact in the near future. These include using countermeasures for code injection vulnerabilities in the C and C++ programming languages, using lightweight specifications for improving software quality, using aspect orientation for security, and threat modeling.

“Winning the Trustworthy Computing RFP award is very important to us, because it signals that the world’s largest software company – one that takes security very seriously from the perspectives of product development and research –believes we’re on the right track in regard to educating students on secure software system design,” says Piessens.

The Trustworthy Computing RFP program encourages academic research focused on the development of Trustworthy Computing curriculum at academic institutions worldwide. This year’s program is the second in a series of Trustworthy Computing and Software Engineering Curriculum RFPs. It focused on advancing Trustworthy Computing by developing innovative technology and policy in five areas: security, privacy, reliability, business practices or secure software engineering. Microsoft has made significant investments in Trustworthy Computing, through collaboration with governments, industrial partners, consortiums and policy boards, as well as academic researchers.

Microsoft Research received more than 112 proposals for the Trustworthy Computing RFP program this year. Of the 15 winners, three are from Belgium, South Korea and Russia, and 26 percent are women.

Alfred Weaver, another RFP recipient, is looking to introduce a Trustworthy Computing course at the University of Virginia in Charlottesville that addresses how to best design healthcare-specific applications that have robust system security. Weaver and his computer-science research team have been conducting cross-disciplinary research with the university’s medical faculty for over a decade. For the past four years, their joint efforts have centered on the relatively new HIPAA requirements: How can computers enforce the privacy and security of protected resources such as the electronic healthcare record? How can administrators express privacy and security policies that computers can implement? How can healthcare professional be reliably identified via remote access?

Weaver says that the academic research community must produce durable answers to questions like these if the PITAC’s healthcare recommendations for embracing electronic patient records and digital clinical tools are to be adopted.

“Until we teach the principles and applications of robust system security at the university level, there will be an insufficient cadre of security-conscious software professionals who can implement the requirements, and thus help reap the benefits, of computer applications that support modern medicine,” says Weaver. “Winning the Trustworthy Computing RFP award gives us an opportunity to collaborate with Microsoft Research on this very important issue.

More RFP Recipients to be Announced

This academic funding is part of Microsoft External Research & Programs group’s broader collaborative model – one that underscores Microsoft’s ongoing commitment to invest deeply in innovative research. The group supports basic research and curriculum innovation in many areas of computing, including social computing and gaming.

“Microsoft is deeply committed to partnering with academia and governments to advance computer science education, cultivate next-generation academic thought leaders, and build knowledge-based economies,” says Sailesh Chutani, director of the External Research & Programs group within Microsoft Research.

In addition to the Trustworthy Computing and Virtual Earth RFP programs, External Research & Programs will soon announce the winners of its $1.2 million Digital Inclusion RFP program, which empowers academic researchers worldwide to tackle technological challenges to positively affect health, education and socioeconomic conditions.

Other recent funding programs have included Computer Gaming Production Curriculum, ConferenceXP, PHOENIX, Table PC and Computing Curriculum, and Excellence in Software Engineering.

“Our goal is to fund innovative and exciting research in the academic community, because a thriving research community ensures the future of the computing industry overall,” says Chutani.

For a complete list of the Trustworthy Computing RFP winners and descriptions of their research projects, visit