Department of Computer Science and Electrical Engineering

Computer Science and Electrical Engineering
University of Maryland, Baltimore County
Ph.D. Dissertation Defense

A Rapidly Deployable Image Classification System Using Feature Views

Adrian Rosebrock

9:00am Friday, 18 April 2014, ITE 346, UMBC

Constructing an image classification system using strong, local invariant descriptors is both time consuming and tedious, requiring much experimentation and parameter tunings to obtain an adequate performing model. Furthermore, training a system in a given domain and then migrating the model to a separate domain will likely yield poor performance. As the recent Boston Marathon attacks demonstrated, large, unstructured image databases from traffic cameras, security systems, law enforcement officials, and citizens can be quickly amassed for authorities to review; however, reviewing each and every image is an expensive undertaking, in terms of both time and human effort. Inherently, reviewing crime scene images is a classification task. For example, authorities may want to know if a given image contains a suspect, a suspicious package, or if there are injured people in the photo. Given an emergency situation, these classifications will be needed as quickly and accurately as possible. In this work we present a rapidly deployable image classification system using “feature views”, where each view consists of a set of weak, global features. These weak global descriptors are computationally simple to extract, intuitive to understand, and require substantially less parameter tuning than their local invariant counterparts. We demonstrate that by combining weak features with ensemble methods we are able to outperform current state-of-the-art methods or achieve comparable accuracy with much less effort and domain knowledge. We then provide both theoretical and empirical justifications for our ensemble framework that can be used to construct rapidly deployable image classification systems called “Ecosembles”.

Finally, we recognize the fact that image datasets give us the relatively unique opportunity to extract multiple feature representations through the use of various descriptors. In situations where the original dataset is not available for further feature extraction or in cases where multiple feature views are ambiguous (such as predicting income based on geographical location and census data) the Ecosemble method cannot be applied. In order to extend Ecosembles to arbitrary datasets of diverse modalities, we introduce artificial feature views using kernel approximations. These artificial feature views are constructed from a single representation of the data, alleviating the need to explicitly extract multiple feature views. We then apply artificial feature views to a diverse range of non-image classification datasets to demonstrate our method is applicable to multiple modalities, while still outperforming current state-of-the-art methods.

Committee: Drs. Tim Oates (chair), Jesus Caban, Tim Finin, Charles Nicholas, Jian Chen

Computer Science and Electrical Engineering
University of Maryland, Baltimore County

Making Physical Inferences to Enhance Wireless Security

Prof. Jie Yang, Oakland University

1:00pm Tuesday, 8 April 2014, ITE 325b

The ubiquity of wireless is redefining security challenges as the increasingly pervasive wireless networks make it easier to conduct attacks for new and rapidly evolving adversaries. There is an urgent need to seek security solutions that can be built into any wireless network stack to defend against attacks across the current heterogeneous mix of wireless technologies, which do not require extensive customization on wireless devices and cannot be undermined easily even when nodes are compromised. In particular, security solutions that are generic across all wireless technologies and can complement conventional security methods must be devised. My research efforts are centered around exploiting physical properties correlated with pervasive wireless environments to enhance wireless security and make inferences for context-aware applications. In this talk, I will present my research work in exploiting spatial correlation as a unique physical property inherited from any wireless device to address identity-based attacks including both spoofing and Sybil. These attacks are especially harmful as the claimed identity of a wireless device is often considered as an important first step in an adversary’s attempt to launch a variety of attacks in different network layers.

Our proposed techniques address several challenges include (1) detecting identity-based attacks in challenging mobile environments, (2) determining the number of attackers, and (3) localizing multiple adversaries. I will also present our work in secret key generation for facilitating secure data communication in the increasing dynamic wireless environments. Our work addressed the problem of collaborative secret key extraction for a group of wireless devices without relying on a key distribution infrastructure. Moreover, in order to provide efficient secret key generation, we exploit fine-grained physical layer information, such as the channel state information made available from OFDM system, to improve the secret key generation rate and make the secret key extraction approach more practical.

Jie Yang received his Ph.D. degree in Computer Engineering from Stevens Institute of Technology in 2011. He is currently an assistant professor in the Department of Computer Science and Engineering at Oakland University. His research interests include cyber security and privacy, and mobile and pervasive computing, with an emphasis on network security, smartphone security and applications, security in cognitive radio and smart grid, location systems and vehicular applications. His research is supported by National Science Foundation and Army Research Office. He is the recipient of the Best Paper Runner-up Award from IEEE Conference on Communications and Network Security 2013 and the Best Paper Award from ACM MobiCom 2011. His research has received wide press coverage including MIT Technology Review, The Wall Street Journal, NPR, CNET News, and Yahoo News.

Hosts: Nilanjan Banerjee and Tim Finin

The UMBC Graduate Cybersecurity Program is offering the following courses over the Summer 2014 session.

Each class will meet one or two days a week in the late afternoon or evening, depending on the length of the session where the course is offered.

• CYBR 620: Introduction to Cybersecurity (Parr)
  T/TH 6-8:45PM
  Summer I (6 weeks) May 27th-July 3rd, 2014
• CYBR 691:  Special Topics in Cybersecurity: “Software Security” (Coman)
  M/W 7:10-9:40PM
  Summer I (6 weeks) May 27th-July 3rd, 2014
  Pre-Req: CYBR 620 or equivalent

Offered to students enrolled at the Universities at Shady Grove campus ONLY

• CYBR 691: Special Topics in Cybersecurity: “Cybersecurity Risk Management” (Shariati)
  M/W 7:10-9:40PM
  Summer I (6 weeks) May 27th-July 3rd, 2014
  PreReq: CYBR 620 and enrollment at the Universities at Shady Grove

The deadline to apply for Fall 2014 admission to the UMBC Graduate Cybersecurity Program is August 1, 2014.

Three students in UMBC’s College of Engineering and Information Technology (COEIT) Department of Computer Science and Electrical Engineering (CSEE) have been selected for major scholarships to pursue studies in cybersecurity-related fields under UMBC’s participation in the National Science Foundation’s (NSF) Scholarship for Service (SFS) Federal CyberCorps program. As SFS Scholars, students receive full tuition, fees, annual reimbursement of professional development expenses ($3,000), a nine-month stipend ($20,000 for undergraduates, $25,000 for MS/MPS students, and $30,000 for PhD students) for up to two years (three years for PhD or BS/MS), and assistance with federal cybersecurity internships and career placement. To be eligible for SFS funding, applicants must be full-time students at UMBC committed to cybersecurity-related research and education activities, have an excellent academic record, and not have any significant outside employment obligations.

The first-round awardees for AY14-15 are:

• Anastasia Raffucci, BS (CMPE)
• Jackson Schmandt, Ph.D. (CMPE)
• Brooke Young, BS (CMSC)

These new awardees will join existing UMBC SFS Scholars Oliver Kubik (BS, CMSC), Mary Mathews (PhD, CMSC), Nathan Price, MS (CMPE). Punlada Muangrat BS (IS), Alex Cooke BS (IS), and Denis Danilin MS (IS).

ADDITIONAL SLOTS AVAILABLE! Applications for SFS support for AY14-15 will be accepted through 12PM (noon) on Friday, May 2, 2014 with notice of awards made by May 23, 2014. All levels of study are welcome to apply — however, students must be admitted to (or already enrolled at) UMBC first before applying for SFS. For information about SFS (including application information and instructions) and other US government-related cybersecurity scholarships available, please visit the CISA website.

The CyberCorps program produces highly-qualified professionals to meet the United States government’s increasing need to protect American’s cyber infrastructure. While in the program at UMBC, SFS CyberCorps Scholars participate in special SFS program activities, have opportunities to engage in mentored research opportunities both at UMBC and its partners from industry and government, and must complete a paid summer internship for the federal government. Upon graduation, each student must work for the government (for pay) for one year for each year of scholarship received. CSEE Drs. Alan Sherman and Richard Forno direct the program under a five-year $2.5 million NSF grant received in 2012.

Arduino Workshops from UMBC commonvision on Vimeo

TL;DR: UMBC’s IEEE student branch will hold free weekly workshops on Arduino starting next week. Register at http://bit.ly/OVzA1f.

The Arduino micro-controller is a great device for anyone who wants to learn more about technology. It is used in a variety of fields in research and academia as well as by hobbyists. Arduino can be used for projects ranging from quadcopters to thermal cyclers  and even wearable electronics. Here is more information about the Arduino.

The workshops are intended for ANY student from ANY major and only require a basic understanding of basic programming concepts such as if-statements and loops. The majority of the code will be provided.

The Arduino workshops are offered on a weekly basis at two levels: Level 1 for those new to Arduino and Level 2 for students who have completed Level 1 or have a basic familiarity with Arduino. The Level 1 workshop is offered on Tuesday and Wednesday from 7:00pm to 9:00pm starting the week of April 1st. You may select to participate in either our Tuesday lecture or Wednesday lecture. The Level 2 workshop is only offered on Thursday from 7:00pm to 9:00pm starting the week of April 1st as well. For both workshops, you will need to bring your laptop and you must able to download the Arduino IDE.

Register at http://bit.ly/OVzA1f. Seats are limited. Contact Sekar Kulandaivel () with questions about registration, downloading the IDE, or anything else.

The next Council of Computing Majors (CCM) will be at Noon this Wednesday, March 26th, in Math/Psychology 008. One of our group members, Patrick McElvaney, will give a presentation about a new CCM project: using the Raspberry Pi and photo-receptors to analyze water in the Chesapeake watershed. His talk will provide an overview of the installation of sensing equipment, specific details on how the photo-receptors can differentiate between organic and inorganic compounds, and information on how other students can get involved.

The CCM is a student organization for undergraduate computer science and computer engineering majors as well as other students interested in computing. If you are interested in starting your own project, please come and speak with any of the officers after the meeting.

Scalable monitoring and kernel learning for energy grids

Vassilis Kekatos
Department of Electrical and Computer Engineering
University of Minnesota

12:00pm-1:00pm, Thursday, 13 March 2014, ITE 325b, UMBC

The smart grid vision urges for enhanced situational awareness, sustainability, and economics over our energy systems. While meters are being installed throughout the grid, algorithms that can effectively process this big data deluge are now demanded. Aligned to that end, this talk focuses first on scalable grid monitoring. Albeit control centers monitor their local grids independently, deregulation and renewables call for power system state estimation (PSSE) at the interconnection level. To address the complexity and communication challenges involved, a decentralized PSSE framework based on the alternating direction method of multipliers has been developed. Beyond conventional least-squares, our framework can identify outliers and circuit breaker statuses as verified on IEEE grids having thousands of nodes. Electricity market inference is the second theme of this talk. We will first demonstrate how grid topologies can be revealed using only publicly available real-time energy prices. This becomes feasible after recognizing that the price matrix can be factorized as the product of the grid Laplacian times a low-rank and sparse matrix. Leveraging the link between energy markets and the underlying physical grids, we will then cast day-ahead price forecasting as a kernel learning task. Through a novel nuclear norm-based regularization, kernels across pricing nodes and hours are systematically selected. Numerical tests using real data from the Midwest ISO market corroborate the interpretative merits of our schemes.

Dr. Vassilis Kekatos is currently a postdoctoral associate with the ECE Dept. of the University of Minnesota, Minneapolis. He obtained his Ph.D. in Computer Engineering and Science from the University of Patras, Greece, in 2007. In 2009, he received a Marie Curie fellowship. During the summer of 2012, he worked as a consultant for Windlogics Inc. His current interests lie in the areas of signal processing, optimization, and statistical learning towards modernizing our energy systems.

Host: Tulay Adali

UMBC’s intercollegiate cyber competition team (the “CyberDawgs”) are heading to the finals of the Mid-Atlantic Collegiate Cyber Defense Competition (CCDC) on March 26-29 at the Johns Hopkins University Applied Physics Lab – Kossiakoff Center in Laurel, MD!

Nearly 300 students from schools throughout the Mid-Atlantic region competed in several qualification rounds to determine the eight finalist teams. Joining the CyberDawgs at the finals will be teams from Anne Arundel Community College, Liberty University, West Virginia University, Towson University, Radford University, Capitol College, and Millersville University. The winner of the Mid-Atlantic finals will advance to the National CCDC Finals in San Antonio, TX later this spring.

The 2014 MA-CCDC finals scenario challenges teams to defend their networks against a series of escalating cybersecurity attacks occuring during a simulated disaster management situation in Maryland.

The 2014 National CyberWatch Mid-Atlantic Collegiate Cyber Defense Competition (CCDC) presented by National CyberWatch Center is in its ninth year of providing a unique experience for college and university students to test their cybersecurity knowledge and skills in a competitive environment.

The UMBC Council of Computing Majors will hold its first hands-on Raspberry Pi workshop from 2:00-4:00 this Friday, March 7, in ITE240.

The Raspberry Pi is a $35 credit-card-sized, single-board computer that runs a version of Unix. Originally developed for teaching computer programming to children, it is now being used in many useful and exciting applications, from near-space weather balloons to baby monitors to media servers. The possibilities are only limited by your imagination.

The initial workshop will cover the Raspberry Pi, its Raspbian Unix OS, and how to program it using Python for real-world applications. There will be 20 Pi computers for participants to use. The workshop is designed so freshman and non-computer science majors can attend and participate. If you know anyone who would be interested in attending, please send them the link and information!

Space is limited, so sign up to reserve a seat.  Intermediate and advanced workshops will follow later in the semester. See the Pi FAQ for general information on the Pi and Raspbian for information on its operating system.

For more information, contact CCM president Austin Murdock ().