Sina Miran
Curriculum Vitae
Contact
Information
1305 A.V. Williams Building,
University of Maryland,
College Park, MD 20742
E-mail:
smiran@umd.edu
sina.miran@gmail.com
Webpage: www.ece.umd.edu/vsmiran
Research
Interests
statistical signal processing, machine learning, compressed sensing and data analytics with
applications in computer vision, signal/image/text processing
Education
PhD/MSc (GPA: 4/4 via 22 credits)
Fall 2014 - Fall 2018 (expected)
University of Maryland, College Park (UMD), MD, US
Major: Signal Processing, Electrical & Computer Engineering
Adviser: Prof. Behtash Babadi
• Coursework: Convex Optimization (A+), Linear Statistical Models (A), Estimation
& Detection Theory (A), Sparse Statistical Signal Processing (A), Deep Learning
Seminar (A), Random Process (A+), Information Theory (A+), Advanced DSP (A)
PhD/MSc (GPA: 3.91/4 via 28 units)
Fall 2013 - Summer 2014
University of California, Santa Barbara (UCSB), CA, US
(Transferred to UMD)
Major: Signal Processing, Electrical & Computer Engineering
• Selected Coursework: Pattern Recognition (A), Advanced Data Mining (A-), Non
Parametric Regression and Classification (A), Stochastic Process (A), Optimal Filtering
and Estimation (A), Digital Image Processing (A-)
Bachelor of Science (GPA: 17.78/20 via 162 units)
Fall 2008 - Spring 2013
Sharif University of Technology (SUT), Tehran, Iran
Major: Communication Systems, Electrical Engineering
• Selected Coursework: Digital Image Processing, Digital Signal Processing, Digital
Signal Processing 2, Digital Signal Processing Lab, Numerical Optimization
Minor: Economics
• Coursework: Game Theory, Microeconomics Intro., Macroeconomics Intro., Econometrics Intro., Principles of Economics, Engineering Economics, International Trade
Pre-University Diploma (GPA: 19.79/20)
Fall 2004 - Spring 2008
Allame Helli NODET High School, Tehran, Iran
Major: Mathematics & Physics
* NODET stands for National Organization for Development of Exceptional Talents
Work
Experience
Research Assistant, ECE Department, UMD, MD, US
• under supervision of Prof. Behtash Babadi
Spring 2016
TA Training and Development Fellow, ECE Department, UMD, MD, US
• Presenting TA workshops and mentoring new TAs
Fall 2015, Spring 2016
Teaching Assistant, ECE Department, UMD, MD, US
• Random Process (ENEE 620) (grad-level)
• Digital Signal Processing (ENEE 425)
• Engineering Probability (ENEE 324)
Fall 2015
Spring 2015
Fall 2014
Teaching Assistant, ECE Department, UCSB, CA, US
• Probability and Statistics (ECE 139)
• Introduction to Computer Vision (ECE 181B)
Spring 2014
Winter 2014
Research Assistant, ECE Department, UCSB, CA, US
• Vision Research Lab (VRL)
Fall 2013
• Working on data acquisition using Kinect cameras for a face recognition application
Teaching Assistant, EE Department, SUT, Tehran, Iran
• Digital Signal Processing Laboratory
1 of 3
Summer 2012
• Digital Signal Processing Course
Spring 2011, Fall 2011, Spring 2012
* My responsibilities as Teaching Assistant included holding discussion sections for recitation
and problem solving, holding office hours, designing parts of HWs and exams, and grading.
Also, I’ve been responsible for evaluating the programming skills of students (C and
MATLAB) and designing DSP-related MATLAB assignments.
Publications
P. Imany, M. Yazdanpanah, S. Miran, M. Showkatbakhsh, “A Devised Approach to
Optimize Color Space Transformation for Image Compression”, In Telecommunications
Forum (TELFOR), 2012 20th , pp. 1737-1740. IEEE, 2012 [PDF Link]
Selected
Projects
Decoding of Sparse Oscillatory Components in Neural Data
Spring 2015
PhD research, under the supervision of Prof. Babadi, ECE Department, UMD
With recent advances in measurement instruments, recording of the data of single neurons has been
made possible. Under certain conditions, these binary signals often exhibit a sparse oscillatory
behavior. However, so far, only some heuristical methods have been used to extract these
oscillatory components. We have introduced a model-based machine learning method to extract
these oscillatory components from neural data overcoming the difficulties of the heuristical methods.
(preparing for submission!)
Nonnegative Matrix Factorization (NMF)
Fall 2015
Convex Optimization Course Project, ECE Department, UMD
NMF is a non-convex optimization problem which has widespread applications in topic modeling,
clustering, and recommender systems. The idea behind NMF is how to factorize an element-wise
nonnegative matrix as a multiplication of two low-rank element-wise nonnegative matrices. In this
project, we implemented several existing methods to deal with non-convex problems consisting of
alternate minimizations, ADMM, QCQP, and lifting to find such a factorization, and we developed
a new method which has superior performance over these in terms of convergence speed and final
residual error norm. (detailed report to be posted!)
Latent Dirichlet Allocatioin for Topic Modeling
Winter 2014
non-Parametric Regression and Classification Course Project, PSTAT Department, UCSB
Implementing an LDA model for topic modeling in a set of text documents and giving an introductory
presentation on the subject. [Presentation]
Sparse Representation for Face Recognition
Digital Image Processing Course Project, ECE Department, UCSB
Fall 2013
Giving a presentation and doing an extensive implementation of algorithms for sparse representation
models in face/object recognition. [Report] [Poster]
PPG Signal Features in Heart Disease Patients
Spring 2013 - Summer 2013
B.Sc. Thesis, under the supervision of Prof. Zahedi, EE Department, SUT
Photoplethysmogram (PPG) signal represents the volumetric changes in blood vessels. The goal of
this project was to do supervised learning by implementing feature extraction (time and frequency
domains), feature selection, and classification methods on the PPG signals of test subjects in order
to see whether subjects can accurately be classified based on the extent of their disease using the
PPG signal features. The correct extent of the disease was determined by angiography, which is an
invasive, costly, and time-consuming method compared to PPG signal acquisition and processing.
[Report]
Color Space Transformation for Image Compression
EE Department, SUT
Summer 2012
Standard JPEG-based color image compression methods utilize predefined color space transformations like RGB to YCbCr to concentrate most of the energy of the image in one plane. In this
project, a new algorithm is proposed to optimize the transformation for each image which resulted
in a paper published in TELFOR 2012. [PDF Link]
Awards &
Honors
Selected as a departmental distinguished TA, ECE department, UMD, 2015
Awarded the Clark School of Engineering Distinguished Graduate Fellowship to study at
the ECE department, UMD, 2014
2 of 3
Ranked in the top %10 in terms of total GPA, entering class of 2008, EE department at
SUT, 2013
Awarded the four-year William Larimer Mellon Fellowship to study at Tepper School of
Business, Carnegie Mellon University, 2013 (Declined)
Awarded the full-year Lord Dahrendorf Scholarship to study at London School of Economics
(LSE), 2013 (Declined)
Ranked 1st in the minor of economics program, Graduate School of Management and
Economics (GSME), SUT, class of 2013
Ranked 122nd in Iranian Nationwide University Entrance Exam (Concours) among more
than 250,000 participants, 2008
Computer
Skills
Programming Skills
• MATLAB, C/C++, Python, R, 8051/8085 Assembly, Verilog
Technical Software
• Visual Studio, Mathcad, Code Composer Studio (DSP boards application), Codevision
AVR, Orcad PSpice and HSpice
References
Available upon request
3 of 3