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

Meet Cheli. She is a Chemical Engineering major and her research explores the microscopic world of Filamentous fungi. She hopes to hypothesize that hyphal branching is, in part, regulated by autophagy; a cellular-level recycling process that occurs in a wide range of species from fungi to humans.

How did you find your mentor for year research project?
During my first semester at UMBC I received an email from my ENES 101 instructor saying that Dr. Marten was looking for new undergraduate researchers. I sent him my resume and high school transcript (I did not yet have a UMBC transcript), and I was accepted a couple weeks later, and have been there ever since.

How much time do you put into it?
I generally put in 12-15 hours a week for research but it varies depending on the experiment I am conducting.

How did you hear about the Undergraduate Research Award (URA) program?
My predecessor in the lab was a URA scholar, so my research mentor suggested I should apply as well.

Was the application difficult to do?
The application was straightforward. I would say that the most difficult part was making sure it did not exceed the page limit.

How much did your mentor help you with the application?
My mentor was the last person to review my application before it was submitted.

What has been the hardest part about your research?
The hardest part about my research has been trying to understand unexpected results. Often, in research, you can reach a road block and the difficult part is trying to figure out if this is the result of human error, an error in theory, or an error in the hypothesis.

What was the most unexpected thing?
Although I was told communication is a key component of research, I was surprised at how much time I spend using this skill. The importance of spending time to communicate results has become increasingly clear, whether it is explaining individual experiments in lab notebooks and weekly reports or developing larger-scale presentations for team members or conferences.

How does your research relate to your work in other classes?
As a chemical engineering major, I have to take classes in a wide variety of fields, and the work inside the lab has really helped me understand the importance of interdisciplinary studies. Genetics, Cell Biology, Statistics and Organic chemistry are all directly related to my field of research, and help me to fully understand the theory behind my research. On the other hand, the work I have done in the lab has given me skills in experimental design (a crucial component of ENCH 225) that I would not have had otherwise.

What is your advice to other students about getting involved in research?
Find your personal motivation to do research and talk to people who are currently involved in your field of interest. Once you have done that, make sure you can find the time in your schedule. This is often the most difficult thing to do, but with the right motivation, it is usually possible. Find research that gets you excited and apply to work in that laboratory. It is always possible to find time to devote to doing something you love.

Read more about her research here.

UMBC Relationship Violence Prevention Advocates

Position Description (Fall, 2014)

(applications  accepted on a rolling basis from April 18 through August 8)

The UMBC Relationship Violence Prevention Advocates program is a comprehensive program funded by a grant from Verizon.  The program changes each year and is designed to educate the UMBC Community on specific relationship violence prevention information.  The goals of the program are to systematically raise awareness and support for a campus to be without relationship violence and create an environment that encourages healthy relationships and the victims of relationship violence to come forward and receive support. Specific components of the program include:

·         Relationship violence prevention education and training for volunteer students, faculty and staff advocates who represent various groups and populations on campus. These advocates will help to educate the campus community about relationship violence prevention through their networks,  and serve as referral sources for those who have experienced or are impacted by relationship abuse/violence or have questions about unhealthy relationships.

·         Training for UMBC athletic coaches

·         Relationship Skills Program via the UMBC Counseling Center

·         Awareness, Education and Resources through www.umbc.edu/rvap

·         Bystander Intervention training through the Green Dot program (VP Student Affairs Office)

General Description:

UMBC Relationship Violence Prevention Advocates are UMBC community members who: a)  are able to provide relationship/dating/intimate partner violence prevention information to their community networks, b) observe the UMBC community and take notice of and report  community issues related to relationship violence to advocate trainers, c) encourage bystander intervention techniques within UMBC community network, d) make appropriate referrals, e) provide helpful resource information to their UMBC community network and individuals, f) support violence prevention efforts on campus, and g) volunteer to staff events (tabling, etc).

1.      Must attend training on August 25, 2014 from 9:30am-4pm

2.      Must attend the following Friday meetings at noon- August 29; September 5, 12, 19; October 3, 17, 31; November 14; December 5

3.      Must share educational information with your network of friends, colleagues, student organizations, team members, and at campus events

4.      Must utilize skills to make referrals, provide resources, intervene safely when necessary using bystander intervention skills.

UMBC’s ACM student chapter invites nominations from the graduate students in the CSEE department for student officer positions for academic year 2014 – 2015 (Fall 2014 to Spring 2015).

ACM (the Association for Computing Machinery) is a premier organization that promotes computing and technology around the US and the world. On the campus, the ACM student chapter is affiliated and supported by the UMBC graduate students association. The goal of the ACM student chapter is to foster interaction between all students, both graduate and undergraduate, in the CSEE department, provide a forum for student interaction, and opportunities for members to expand their knowledge of computing.

The positions available (and their general responsibilities):

These positions are open to graduate students only. If elected, you would be required to signup as an ACM student member. Membership fee is $19 only.

Please email us the position you would like to run for (there will be elections if we get multiple nominations for a position). Alternatively you can nominate any other person for the positions above. In that case, please send their name, email address and which position you would like to nominate them for.

Please send in your nominations by end of day, Monday, April 14, 2014 to acmofficers at lists dot umbc dot edu. Elections will take place the following week (venue, date and time to be announced later).

UMBC’s ACM student chapter invites nominations from the graduate students in the CSEE department for student officer positions for academic year 2014 – 2015 (Fall 2014 to Spring 2015).

ACM (the Association for Computing Machinery) is a premier organization that promotes computing and technology around the US and the world. On the campus, the ACM student chapter is affiliated and supported by the UMBC graduate students association. The goal of the ACM student chapter is to foster interaction between all students, both graduate and undergraduate, in the CSEE department, provide a forum for student interaction, and opportunities for members to expand their knowledge of computing.

The positions available (and their general responsibilities):

These positions are open to graduate students only. If elected, you would be required to signup as an ACM student member. Membership fee is $19 only.

Please email us the position you would like to run for (there will be elections if we get multiple nominations for a position). Alternatively you can nominate any other person for the positions above. In that case, please send their name, email address and which position you would like to nominate them for.

Please send in your nominations by end of day, Monday, April 14, 2014 to acmofficers at lists dot umbc dot edu. Elections will take place the following week (venue, date and time to be announced later).