Signal Processing Assistant
Resume:
Nevena Lazic
******@****.********.** Office: 1-416-***-****
www.apl.utoronto.ca/people/nevena Mobile: 1-416-***-****
Edward S. Rogers Sr. Dept. of Electrical and Computer Engineering
University of Toronto, 10 King s College Rd., Toronto ON, M5S 3G4, Canada
EDUCATION
Ph.D., Electrical Engineering 2005-2011
University of Toronto
Advisors: Brendan J. Frey and Parham Aarabi
Thesis : Message Passing Algorithms for Facility Location Problems
M.A.Sc., Electrical Engineering 2003-2005
University of Toronto
Advisor: Parham Aarabi
Thesis: Communication Over an Acoustic Channel Using Data Hiding Techniques
B.A.Sc., Computer Engineering 1999-2003
University of Toronto
GPA: 3.91/4
GRADUATE LEVEL COURSES
Stochastic Processes, Information Theory, Error Correcting Codes, Introduction to
Cryptography, Data Communication, Machine Learning, Graphical Models,
Mathematical Statistics, Convex Optimization
AWARDS AND SCHOLARSHIPS
Ewing Rae Graduate Scholarship 2010
Viola Carless Smith Research Fellowship 2009
NSERC Postgraduate Scholarship D 2005-2007
NSERC Postgraduate Scholarship M 2003-2005
U of Toronto Dean s Merit Scholarship (declined) 2003
NSERC Undergraduate Student Research Award 2002
U of Toronto Dept. of ECE Outstanding Student Award 2001-2002
U of Toronto Scholar Award 2001-2002
John M. Empey Scholarship 2000-2001
Baptie Scholarship 2000-2001
U of Toronto Scholar Award 1999-2000
U of Toronto Dept. of ECE Admission Scholarship 1999-2000
RESEARCH INTERESTS
Developing principled approaches to solve practical problems in machine
learning, computer vision and computational biology.
RESEARCH EXPERIENCE
Research Assistant 2005-2011
University of Toronto, Probabilistic Statistical Inference (PSI) Group
Advisors: Brendan J. Frey and Parham Aarabi
Worked on exemplar-based clustering via maximum-a-posteriori (MAP)
inference. Extended and generalized the graphical model underlying the Affin ity
Propagation clustering algorithm.
Constructed a new 3-approximation algorithm for the Uncapacitated Facili ty
Location problem using message passing inference in a graphical model.
Devised a novel approach to 3-D rigid body motion segmentation from point
correspondences, formulating the problem as MAP inference.
Summer Research Intern Summer 2010
Microsoft Research, Redmond WA
Advisor: Nebojsa Jojic
Worked on modeling the evolution of HIV in response to pressure from the host
immune system.
Summer Research Assistant Summer 2009
University of Toronto, Identity, Privacy and Security Institut e (IPSI)
Surveyed material for ECE1517 Biometric Systems, a graduate -level course.
Created lecture notes and presentation slides.
Research Assistant 2003-present
University of Toronto, Artificial Perception Laboratory (APL)
Advisor: Parham Aarabi
Developed a novel system for low-rate acoustic data communication, using
digital audio watermarking techniques to embed bits into music.
Explored the properties and performance of low-level image descriptors on a
number of tasks in computer vision and speech detection.
NSERC Undergraduate Student Researcher Summer 2002
University of Toronto, Artificial Perception Laboratory
Advisor: Parham Aarabi
Developed and implemented a novel system for automatic robot navigation,
using microphone array based sound local ization.
PUBLICATIONS
Solving the Uncapacitated Facility Location Problem Using Message Passing
Algorithms.
N. Lazic, B.J. Frey and P. Aarabi. In Proceedings of the 13th International Conference on
Artificial Intelligence and Statistics (AISTATS), 2010.
FLoSS: Facility Location for Subspace Segmentation
N. Lazic, I.E. Givoni, B.J. Frey and P. Aarabi. In Proceedings of the 12th IEEE International
Conference on Computer Vision (ICCV), 2009.
Spoken Term Detection Using Visual Spectrogram Matching
N. Lazic and P. Aarabi. In Proceedings of the 10th IEEE International Symposium on
Multimedia (ISM), 2008.
Importance of Feature Locations in Bag -of-Words Image Classification
N. Lazic and P. Aarabi. In Proceedings of the 32nd IEEE International Conference on
Acoustics, Speech, and Signal Processing (ICASSP), 2007.
Communication Over an Acoustic Channel Using Data Hiding Techniques
N. Lazic and P. Aarabi. In IEEE Transactions on Multimedia, 2006.
Adaptive Time-Frequency Data Fusion for Speech Enhancement
G. Shi, P. Aarabi and N. Lazic. In Proceedings of the 6th International Conference on
Information Fusion, 2003.
INVITED TALKS AND PRESENTATIONS
Solving the Uncapacitated Facility Location Problem Using Message Passing
Algorithms
Workshop for Women in Machine Learning ( WiML), Vancouver B.C., Dec. 2010.
13th International Conference on Artificial Intelligence and Statistics (AISTATS),
Sardinia, May 2010. (http://videolectures.net/aistats2010_lazic_stufl/ )
Machine Learning Seminar, University of Toronto, Oct. 2009.
FLoSS: Facility Location for Subspace Segmentation
Workshop for Women in Machine Learning (WiML), Vancouver B.C., Dec. 2009.
Importance of Feature Locations in Bag -of-Words Image Classification
32nd IEEE International Conference on Acoustics, Speec h, and Signal Processing
(ICASSP), 2007.
Sound Localization Using Microphone Arrays
Interactive demonstration in the Artificial Perception Laboratory, as a part of Fall
Campus Day tours, University of Toronto, Sept. 2005.
Audio Data Hiding and the Human Auditory System
Guest lecture Microphone Arrays and Applications (graduate -level course),
University of Toronto, Nov. 2004.
PROFESSIONAL ACTIVITIES
Reviewer
European Conference on Computer Vision (ECCV), IEEE Signal Processing Letters
(SPL), IEEE Transactions on Multimedia (TMM)
Executive Board Member
Annual workshop for Women in Machine Learning (WiML)
TEACHING EXPERIENCE
Teaching Assistant 2003-2010
University of Toronto, Dept. of Electrical and Computer Engineering
MAT290 Advanced Engi neering Math
ECE216 Signals and Systems
ECE253 Digital and Computer Systems
ECE302 Probability and Random Processes
ECE316 Communication Systems
ECE431 Digital Signal Processing
WORK EXPERIENCE
Summer Intern Summer 2001
Bioscrypt Inc., Toronto ON
Reviewed and summarized coding guidelines for the Texas Instruments
eXpressDSP Algorithm Standard, developed sample C c ode.
TECHNICAL SKILLS
MATLAB, C, LaTeX, Assembly, Verilog
Contact this candidate