projects
:: Purpose-Driven
Navigation: Navigation in the context of digital maps is associated
with a sequence of pan and zoom operations that lead to a specific destination.
In this work, we propose creating rich navigational schemes by augmenting the existing
concept of navigation with knowledge of purpose behind it. The proposed technique
enables support of navigational interactions like ``scan region'' and ``explore
neighborhood''.
> People involved: Neeharika
Adabala, Kentaro Toyama
:: Integration of
Maps and Blogs: Location information figures prominently
in one’s identity and daily life. In this work, users are encouraged to create lists
of their personally important locations, which can be placed on their blogs with
two-way links to maps. A demo implements this as an integration of Virtual Earth
and MSN Spaces.
> People involved: Sean Blagsvedt
> Links: http://viavirtualearth.com/vve/Gallery/Default.ashx; http://loblog.viavirtualearth.com/
:: Project Lachesis: Analyzing Location Histories: Advances in location-aware devices allow people to easily
collect their own location history – a record of a subject’s location over some
time. (Location histories are often called “tracks” or “breadcrumbs” by GPS enthusiasts.)
Project Lachesis seeks to model and mine this data for rich semantic content
>
People involved: Kentaro Toyama,
Rahul Gupta, Ramaswamy Hariharan, Yamini Kannan
> Links: http://research.microsoft.com/~toyama/lachesis.htm
::
Text-Free UIs for Maps: Maps are powerful tools
for navigation, but they are highly dependent on text and the ability to read. Using
methods of ethnographic design, we are working on map interfaces that can be used
by illiterate or semi-literate users.
>
People involved: Kentaro Toyama, Aman Sagar
:: Woodcut Maps: We are developing
methods to take a vector representation of map features as input and to render a
map in the style of hand-created maps, such as antique woodcut maps. The patterns
are created with procedural approaches that depend on the semantic labeling on each
vector element, with, for example, mountain ranges drawn differently from coastlines.
> People involved: Neeharika Adabala, Kentaro
Toyama
> Links: A poster paper was presented at SIGGRAPH 2005: http://www.siggraph.org/s2005/main.php?f=conference&p=posters |
 |
:: WWMX: World-Wide Media eXchange: The combination of
photos with location information is powerful: maps can be enhanced by photographs
of visible landmarks, and photos are placed in their geographic context. This project
explored the possibilities of a large database of photos, all of which were annotated
by the latitude and longitude where they were shot.
> People involved: Kentaro Toyama, Ron Logan,
Asta Roseway
> Links: The project is online at http://wwmx.org
:: Aerial Image Analysis: The success
of Visual Earth has brought into focus the potential and impact of well annotated
aerial images. Unfortunately, while access to aerial imagery is becoming cheaper
by the day, the cost of manually annotating this ever increasing amount of data
is rising just as quickly. In this project, we aim to leverage computer vision and
machine learning techniques to automatically annotate regions of interest in aerial
images.
> People involved: Manik Varma, Kentaro Toyama
:: Reading Image Text: Most of us, at some point in
our lives, have been frustrated by our inability to read text in a foreign language
-- "Is this the bus that I should catch"? This project aims to alleviate such frustration
by building a vision algorithm which will recognise text in low resolution images
of real world cluttered scenes (such as those taken by typical mobile devices).
In addition, a text recognition algorithm applied to "location aware" images, such
as those in the WWMX project, would find many interesting applications in map search
as well as location detection.
>
People involved: Manik Varma, Kentaro Toyama, A.
Kumaran and Rajesh Veeraraghavan |
