-
Head tracking by AC-magnetic tracker using adaptive extended Kalman filter
 Head tracking is widely used in human-computer interaction systems and virtual reality environments.
The perceived latency between head motion and computer response causes loss of immersion for the user which can result in dizziness in extreme cases.
Additionally, noise in the head orientation data can result in a ¡°swimming effect¡± that destroys the illusion of virtual reality.
The Extended Kalman Filter (EKF) is typically used to compensate for system latency and can provide optimal estimation if the filter
is correctly tuned to the incoming data. In head tracking applications, the large dynamic range of head motion and the unknown noise
content of the incoming data require a compromise of the tuning parameters that results in less than optimum performance.
Additionally, prediction amplifies the output error of the filter when the state vector is projected out in time.
Adaptive filter techniques can be applied to the EKF to dynamically mitigate the undesirable behavior of the filter due to less than optimum tuning.
Head tracking Kalman filters typically use a constant velocity model that models changes in head motion as random noise.
Constant velocity filters work well for slow head motion but suffer from significant overshoot during sudden movement,
especially if large prediction times are used. The project will investigate other system models such as constant acceleration in an attempt
to resolve the overshoot issue.
Doctoral Student: Henry Himberg
Support: Polhemus Inc educational scholarship
Faculty: Yuichi Motai, School of Engineering (UVM)
-
Vehicular collision avoidance system by inter-vehicle communications
 This system will use GPS mapping data to know what is ahead and determine intelligently
if there could be a collision ahead. It needs to be accurate and it should not give many false warnings, especially in curves, like it is happening with the
systems designed in other research papers. Each car will communicate to other nearby cars its location, direction, and speed. Each car will process the information
received from nearby cars and calculate if the paths will meet ahead, and if so it will give the driver a warning. The system will judge collisions based on speed
and road ahead and will give a warning if close to minimum breaking distance to prevent accident. Communication between cars could be done either via ad-hoc WAN
or Bluetooth.
Doctoral Student: Cesar Barrios
Support: IBM educational scholarship
Faculty: Adel Sadek and Yuichi Motai, School of Engineering (UVM)
-
Learning by imitation for behavioral motions between multiple humanoid robotics
 Humanoid Robots are highly configurable automated devices used in a variety of applications. Previous work has been done to mimic the motion of a human or another robot. This is often done using a visual sensing camera attached to the same processor that the robot is controlled by. We hope to explore the area of continuous repetition. This will involve two robots both possessing a camera and controlling processor. The robots will then take turns repeating the actions of the other robot. This will mimic what would happen if a long series of robots continuously repeated the actions of the robot before it. Another application of this would be an environment where many robots repeated the action of a single robot. Part of this work will involve a system to detect motion of the other robot and extrapolate the individual humanoid robot servo motor position data. At the end of the experiment we will measure the quality of repetition by reading out the position of the final robot¡¯s servo motors and comparing them with the initial controlled robot¡¯s position.
Master Student: Scott Vento and Danniel Couture
Support: IBM educational scholarship, NSF EPSCoR
Faculty: Yuichi Motai, School of Engineering (UVM)
-
An incremental on-line framework of multiple support vector machine, kernel principal component analysis, and kernel feature
analysis for behavioral classification of articulated motions
 When the dimension of input data is very low, sometimes it's very hard to linearly
classify the data. The higher dimension will lead to better classification performance, but meanwhile the computation complexity is also increased. Kernel
method is proposed to use the advantage of higher dimension while still maintain the reasonable computation time. Our project is to specifically study the
kernel PCA which can be used to reconstruct the data in the feature space, de-noise the input data, or cluster the unlabeled data. The traditional Kernel PCA
needs to solve the eigenvalue problem with the same size as the training data set. In addition, each eigenvector is the combination of all training data, which
makes it unacceptable for on-line learning. We proposed a new method to improve the efficiency of kernel PCA. Experiment results demonstrated that the performance
of the eigenvectors we selected is very close to the kernel PCA, and the computational complexity is much less than the traditional kernel PCA.
Doctoral Student: Xianhua Jiang and Mariette Award
Support: NSF EPSCoR and IBM educational scholarship
Faculty: Xingquan Zhu, Robert Snapp, and Yuichi Motai, College of Engineering and Math (UVM)
-
Rodent behavior identification with near-infrared illumination
 This research project uses computer vision methods to identify behaviors of rodents
such as rats and mice, and to track their position in an enclosed, pre-defined space. Automated position tracking and behavior analysis is
important and useful for the areas of behavioral pharmacology (to determine effects of medication in a pre-clinical setting), genetics (behavioral phenotyping),
and drugs of addiction (mechanisms and treatments). Color and motion based algorithms were developed for position tracking, and decision trees were used for
classifying behaviors based on various attributes including area (from an overhead 2D view), perimeter, circularity, etc. Current work involves refining
methods of identifying behavior, and learning the behaviors using ID5R (incremental decision tree).
Doctoral Student: J. Brooks Zurn
Support: MED Associates Inc, NSF EPSCoR
Faculty: Steven Dworkin, Department of Psychology (NCU) and Yuichi Motai, School of Engineering (UVM)
-
Distributed camera networks for constructing a three-dimensional model and target detection
 Distributed camera networks require coordination among multiple cameras.
Given a number of cameras observing the same scene from different viewpoints, the aim is to synthesis a three-dimensional view of the scene.
Once the three-dimensional scene is constructed, it will be used as a background in which moving targets can be detected. Out goal is threefold:
first, construct a three-dimensional view from mobile and autonomous cameras. Second, develop a method for detecting targets within the constructed view,
and last, perform the former two operations under the constraints of low-data rate-- reducing data rate in distributed camera networks has become a
necessary requirement as heterogeneous distributed sensory networks have evolved.
Doctoral Student: Lu'ay Bakir
Support: IBM educational scholarship
Faculty: Yuichi Motai, School of Engineering (UVM)
-
Real-Time depth map estimation using far-infrared camera, range finder, and gyroscope for visual driving assistance
 We are developing unique assistance system for driving a vehicle.
We wish to establish an ideal intelligent vehicle that detects obstructions through use of the mounted sensors, which provide sufficient visual-based aid.
To this end, we utilize an infrared camera, a laser range finder, and a gyro sensor so that these sensors provide some assisted visual cues for a driver
to identify unexpected obstructs. In this proposed vehicle, we develop several interesting modules. For example, in addition to a 2D heat-map provided
by the infrared camera, the system provides depth maps by utilizing the tracking results of the gyro-scope. The depth values in the region of interests
are computed in a real-time video-frame rate by motion stereo from a single far-infrared camera, based on a pinhole lens model. We also develop a graphical
representation of a virtual camera so that the driver can see the field of view presented by the mounted sensor. The laser range sensor will compensate
the camera by increasing the field of view and providing accuracy for detected objects.
Master Student: Michael Finnefrock
Support: NSF EPSCoR, Goodrich educational scholarship
Faculty: Yuichi Motai, School of Engineering (UVM)
-
Mobile robot navigation with a Pan Tilt Zoom camera and an infrared camera
 The research develops an active stereo vision, flexible eyes for a mobile robot.
We focus on intelligent control of robot motion (pan, tilt) and properties (zoom, iris) to extract quick and useful information for a moving robot.
Two Pan Tilt Zoom (PTZ) cameras responds actively to the scene, based on prior knowledge of changes happening from images to images. Moreover,
it is possible to select landmarks in the scence for extra attention or tracking. This will bring greater autonomy to a mobile robot.
The cameras provide the best information for a robot to know about the environment, therefore is able to know its position. However,
the use of cameras requires continuous attention on the sequences of images captured, posing the question of how they can be best processed to
provide localization information. Research has been done in using active stereo vision for the SLAM, localization problem with visual cameras.
However, the combination of visual and infrared images, which can be very useful in environments with heated landmarks, is not adequately addressed.
This project will focus on processing images from an infrared camera and a PTZ camera to localize a mobile robot, as it navigates in an environment with known landmarks.
The position track is maintained under the general probabilistic of Extended Kalman Filter.
Master Student: Hien Nguyen, Paul Montane
Support: NSF EPSCoR, Graduate College at UVM
Faculty: Dryver Huston and Yuichi Motai, School of Engineering (UVM)
|
