Hors D'oeuvres Anyone?
-
Each year, the American Association for Artificial
Intelligence holds a robot competition. This year the
competition has two parts: the Scavenger Hunt, and the
Hors D'Oeuvres Anyone? event.
The Scavenger Hunt would be a new contest for us,
but we should be getting a fast,
computationally-loaded robot this fall that is perfect
for it. The point of the scavenger hunt is to search
for objects in the permissible area (basically a
convention center floor with lots of booths and
obstacles). The capabilities of the new robot should
make it possible to move quickly and recognize objects
in the environment in real time. This will be an
exciting, cutting-edge project that gets you to Austin,
TX next July for the competition.
The Hors D'oeuvres Anyone? competition has robots
serve hors d'oeuvres at a reception. This is a
real-world task with no bounds, human-computer
interaction, lots of noise, and people who try to make
the robot screw up.
Dr. Maxwell has led the winning team for the Hors
D'oeuvres Anyone competition for the past two years,
and would like to take yet another a team of
Swarthmore students back to try again in 2000. Some of
the things that need doing include: 1) improvement of
a the people-finder, 2) improvement upon our algorithm
"working" a crowd, 3) the design and customization of
the robot (which we will purchase this fall) to enable
it to interact with people in a more natural manner,
4) improvement of our person recognition algorithm 5)
The development of an adaptable personality that can
deal with varying conditions like noise, crowds, empty
spaces, and various size rooms.
Note that this project has generated a lot of interest
in the local area and beyond. If you decide to work on
this project, there is a good chance we will demo the
robot during the course of the semester.
-
Any number of students could work on various aspects
of these projects, which will be a good hands-on
introduction to robots, computer vision, and speech
recognition.
|
Building a pan-tilt mount for a stereo camera system
-
A pan-tilt mount is a device that can rotate a load
around two axes. Panning is horizontal rotation, and
tilting is vertical rotation. For an example of a
pan-tilt device, check out http://www.DPerception.com/.
In order for a robot, or other visual human-computer
interface station to have a flexible vision system
that can orient on something of interest, it must have
the ability to rotate. The goal of this project would
be to design a cost-effective pan-tilt head for two
cameras (about a 4 lbs load) that could be controlled
through the serial port of a computer. If successful,
such a device could have commercial potential, as
pan-tilt heads typically cost $2000 and up.
Note: this project could easily be integrated with
the robot projects listed above.
-
One or two students would be appropriate for this
project, which will involve control theory, working
with motors, computer-device communication, and
programming a microcontroller or FPGA.
|
Navigating Hicks with one or more robots
-
Having robots navigate and map buildings is necessary
for them to start to do useful things like collecting
and distributing mail, looking for and finding people,
and other useful tasks (be creative). Navigating Hicks
offers many challenges, including forcing the robot to
be able to find and use the elevator.
in 1999, Carl Wellington and Roger Bock designed and
built an adjustable arm and finger for pushing
elevator buttons. Their project is close to complete,
and the next task is to integrate their arm mechanism
with a vision system and AI system for managing the
process of using an elevator. There is also room for
improvement with the elevator arm, so there is a
mechanical/control theory aspect to this project.
It is likely we could have two or more robots doing
this task, so cooperation is possible. An interesting
variation of this project involves having the robots
try to play hide and go seek, which would be fun to
program and watch.
-
Up to 4-5 students could work on various aspects of
this task, many of which overlap with the other robot
tasks described herein.
|
Developing an advanced segmentation system for content-based
image retrieval
-
One of my major interests is intelligent segmentation
of images: dividing an image up into regions that
correspond closely to the objects in the image. This
is an extremely challenging problem (try defining
"object"), but I have made some good progress over the
past 6 years and want to continue this work.
At its core, this project will involve using your
knowledge of physics, engineering, and mathematics to
devise ways to define and test whether a given
boundary is between two objects of just divides a
single object. This is fundamental computer vision at
its best, so come talk with me if you might be
interested. There are lots of things to explore and
you should be able to build upon my work to develop a
fairly sophisticated approach to the project during
the course of the semester.
-
One or two students would be appropriate for this
project, which will involve most programming and
probably some numerical receipes-type algorithms.
|
Development of a Calibrated Imaging & Simulation Lab
-
In 1999 Ani Hsieh began the development of a
calibrated imaging and simulation lab by designing and
constructing a 6 DOF positioning device capable of
positioning a 2lb load within a 1m workspace. The
next step in this project is to calibrate and test the
device for accuracy, and add additional components
depending upon the results. In addition, we can work
on improving the control algorithms, develop a GUI
interface for the positioner, and use it in
simulations and tracking tasks such as locating and
moving towards people in the lab.
-
One or two people could work on this task, which could
involve GUI development, control theory, mechanical
design, computer vision, or any combination of the
these areas.
|
Robot vision: mapping what the robot sees and identifying
landmarks
-
This is a standard problem in robotics/computer
vision. Typically, roboticists use easy-to-spot
landmarks like bar codes or brightly colored
patterns. For a robot to navigate successfully in a
non-specialized environment, it must be able to find
and remember natural landmarks.
-
One or two students could work on this project, which
will involve mostly programming and testing out
different object recognition/detection
techniques.
|
Computer graphics project: going for realism in textures and
colors
-
There are numerous directions that projects in
computer graphics could take. Dr. Maxwell's personal
interest is in creating realistic scenes using
physics-based models and complex textures and colors,
but your ideas are certainly welcome.
-
Any number of students could work on appropriate
projects in computer graphics. These projects will
include physical modeling of object properties,
physics-based models of light-matter interaction, some
numerical algorithms, and a significant amount of
programming in computer graphics.
|
Pattern recognition & segmentation for automatic typology
generation with hydrogeological data
-
This is a project that Dr. Maxwell is continuing to
work on with a Hydrogeologist in Kansas. The LOICZ
data set is a hydrogeologic data set that contains
data from coastlines around the world. Hydrogeologists
are currently using this data to develop a typology of
coastlines (is Antarctica like New England?).
We are using computer vision/pattern recognition
segmentation and clustering techniques to do a
data-driven classification of the coastlines. There
are multiple computer vision techniques which are
appropriate to this task, and an E90 project could
focus on applying 2-3 of them and comparing the
results. This is a "hot" problem in hydrogeology and
there is a high probability of getting a paper
published from this project.
In 1999 Jennifer Vander Veer developed a GUI for the
LOICZ clustering tools. The hydrogeologists love it,
and now they want version II. Most of this development
will be in the science, rather than the interface, and
you should see results fairly quickly. This project
will also involve developing advanced visualization
techniques in order to more easily assess the data and
the resulting clusters/typology.
-
One or two students would be appropriate for this
task, which will involve programming and visualization
of large scientific data sets.
|
Designing a keypad security system for the Hicks labs
-
Ever forget your key and couldn't get into your lab
when the project is due the next day? This project
would involve putting in a keypad security system for
one or more of the Hicks labs so that students could
have a pass code rather than a physical key.
-
Up to two or three students could work on various
aspects of this project (or different labs), which
will involve digital logic design, and integration
with a microcontroller to run the system
|