ECE Department Calendar

Jun
1
Sun
Census deadline
Jun 1 – Jun 3 all-day
Census deadline
Deadline to apply for graduation
Jun 1 – Jun 3 all-day
Deadline to apply for graduation
Jun
2
Mon
Defense: H. Pourzand
Jun 2 @ 12:00 pm – 2:00 pm

UNIVERSITY OF UTAH
ELECTRICAL AND COMPUTER ENGINEERING DEPARTMENT

THESIS DEFENSE FOR THE DEGREE OF
MASTER OF SCIENCE

by

Hoorad Pourzand
Advisor: Massood Tabib-Azar


Microelectromechanical gyroscopes are readily used in cars and cell phones. Tactical gyroscopes are readily available inexpensively and they offer 0.01 to 0. 1 percent scale factor in accuracy. On the other hand, strategic gyroscopes with much better performance are 100,000 more expensive. The main objective of this work is to explore the possibility of developing in-expensive strategic grades gyroscopes using MEMS.

Most of the available gyroscopes are surface micro machined due to fabrication issues and misalignment problems that are involved in multistep fabrication processes necessary to use the bulk of the wafer as the proof mass in the MEMS gyroscope. It can be shown that the sensitivity of the gyroscope is inversely proportional to the natural frequency; so if bulk micromachining technique is used it is possible to decrease the natural frequency further than current limits in order to increase sensitivity. This thesis is focused on proposing a way to use bulk of the silicon wafer in the gyroscope to decrease the natural frequency to very low levels such as sub kilohertz regime that cannot be achieved by single mask surface micromachining processes and then proposing a solution for solving the misalignment problem caused by using multiple fabrication steps and masks instead of using only one mask in surface micro machined gyroscopes.

In our design discrete proofmasses are linked together around a circle by compliant structures to insure highest effective mass and lowest effective spring constant. By using a proposed double sided fabrication technology the effect of misalignments on frequency mismatch can be reduced. ANSYS simulations show that 20 µm misalignment between the masks causes a frequency shift equal to 0.3% of the natural frequency, that can be compensated using electrostatic frequency tuning. Acceleration parasitic effects can also be a major problem in a low natural frequency gyroscope. In our design a multiple sensing electrode configuration is used that cancels the acceleration effects completely. The sensitivity of the gyroscope with 3126 Hz natural frequency is simulated to be 574 mV/(Deg/sec) or about four times higher than 132 mV/(Deg/sec) , which was used as a benchmark for a sensitive gyroscope.


Monday, June 2, 2014
12:00 p.m.
2610 SMBB

The public is invited

Jun
3
Tue
Defense: Seyedhosseini
Jun 3 @ 10:00 am – 12:00 pm

UNIVERSITY OF UTAH
ELECTRICAL AND COMPUTER ENGINEERING DEPARTMENT

DISSERTATION DEFENSE FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

by

Mojtaba Seyedhosseini
Advisor: Tolga Tasdizen

Scene Labeling with Supervised Contextual Models

Scene labeling is the problem of assigning an object label to each pixel of a given image. It is the primary step towards image understanding and unifies object recognition and image segmentation in a single framework. A perfect scene labeling framework detects and densely labels every region and every object that exists in an image. This task is of substantial importance in a wide range of applications in computer vision. Contextual information plays an important role in scene labeling frameworks. A contextual model utilizes the relationships among the objects in a scene to facilitate object detection and image segmentation. Using contextual information in an effective way is one of the main questions that should be answered in any scene labeling framework.

In this dissertation, we develop two scene labeling frameworks that rely heavily on contextual information to improve the performance over state-of-the-art methods. The first model, called the multi-class multi-scale contextual model (MCMS), uses contextual information from multiple objects and at different scales for learning discriminative models in a supervised setting. The MCMS model incorporates cross-object and interobject information into one probabilistic framework and thus is able to capture geometrical relationships and dependencies among multiple objects in addition to local information from each single object present in an image. The second model, called the contextual hierarchical model (CHM), learns contextual information in a hierarchy for scene labeling. At each level of the hierarchy, a classifier is trained based on downsampled input images and outputs of previous levels. The CHM then incorporates the resulting multi-resolution contextual information into a classifier to segment the input image at original resolution. This training strategy allows for optimization of a joint posterior probability at multiple resolutions through the hierarchy. We demonstrate the performance of CHM on different challenging tasks such as outdoor scene labeling and edge detection in natural images and membrane detection in electron microscopy images.

We also introduce two novel classification methods. WNS-AdaBoost speeds up the training of AdaBoost by providing a compact representation of a training set. Disjunctive normal random forest (DNRF) is an ensemble method that is able to learn complex decision boundaries and achieves low generalization error by optimizing a single objective function for each weak classifier in the ensemble.

Finally, a segmentation framework is introduced that exploits both shape information and regional statistics to segment irregularly shaped intracellular structures such as mitochondria in electron microscopy images.

Tuesday, June 3, 2014
10:00 a.m.
Evans Conference Room
3780 WEB (3rd floor)

The public is invited

Jun
4
Wed
ECE Seminar – Control and Diagnosis of Discrete Event Systems: Some Recent Trends @ Warnock (WEB) 1250
Jun 4 @ 3:00 pm – 4:00 pm
ECE Seminar - Control and Diagnosis of Discrete Event Systems:  Some Recent Trends @ Warnock (WEB) 1250

Stephane Lafortune

University of Michigan-Ann Arbor

When: Wednesday, June 4, 2014 at 3:00 p.m.
Where: Warnock 1250

Abstract

We will present an overview of recent research trends in control and diagnosis of Discrete Event Systems (DES) that are motivated by challenges arising in cyber-physical systems.

In the first part of the talk, we will review the basic theory of supervisory control of DES, discuss its connection with reactive synthesis in computer science, and present results on its application to the problem of collision avoidance in vehicular networks. In this application, the continuous dynamics of the vehicles are abstracted in a discrete-event model, where uncontrollable events capture unmodeled dynamics and unobservable events capture measurement uncertainty.

In the second part of the talk, we will review the basic theory of fault diagnosis in partially-observed DES and then discuss recent work on enforcement of opacity, a class of properties studied in computer security. Opacity is essentially the dual of diagnosability. We will discuss how opacity enforcement techniques can be used to protect users’ privacy in location-based services.

Speaker Biographies

Stéphane Lafortune is a Professor in the Department of Electrical Engineering and Computer Science at the University of Michigan, Ann Arbor. He obtained his degrees from Ecole Polytechnique de Montréal (B.Eng), McGill University (M.Eng), and the University of California at Berkeley (PhD), all in electrical engineering. Dr. Lafortune is a Fellow of the IEEE (1999). His research interests are in discrete event systems and include multiple problem domains: modeling, diagnosis, control, optimization, and applications to computer systems. He co-authored, with C. Cassandras, the textbook Introduction to Discrete Event Systems (2nd Edition, Springer, 2008). He is co-developer of the software packages DESUMA and UMDES.

Jun
12
Thu
First Session Classes: Last day to reverse CR/NC option
Jun 12 – Jun 14 all-day
First Session Classes: Last day to reverse CR/NC option
Jun
17
Tue
First Session Classes: Classes end
Jun 17 – Jun 19 all-day
First Session Classes: Classes end
Jun
18
Wed
Second Session Classes: Classes begin
Jun 18 – Jun 20 all-day
Second Session Classes: Classes begin
Jun
19
Thu
Last day to withdraw from classes
Jun 19 – Jun 21 all-day
Last day to withdraw from classes
Jun
27
Fri
Second Session Classes: Last day to drop (delete) classes
Jun 27 – Jun 29 all-day
Second Session Classes: Last day to drop (delete) classes