Resume

resume_a

This page was updated on 21/10/2018.

Short PDF version of my resume.

Research Interests

  • Deep Learning
  • Machine Learning
  • Computer Vision
  • Computational Genomics
  • Artificial Intelligence
  • Internet of Things

Education

  • 2015-present: M.Sc. Student in Computer Science, Technical University of Munich, Germany
    Cumulative GPA: 1.1 – on 1.0 – 5.0 scale, 1.0 is the highest possible grade.
  • 2011-2015: B.Sc. Student in Computer Engineering – Hardware, University of Isfahan, Iran.
    The admission requirement is to rank top 0.1% in Iran’s college entrance exam
    Cumulative GPA : 18.57/20 (German Scale: 1.4/5) [Top Student: Ranked 1/28]

Research Experiences

  • Software Engineering Intern at Google, Research and Machine Intelligence, May 2018 – September 2018.
    – Researched on Reinforcement Learning for graph-structured data. Attention Networks, TensorFlow
  • Research Assistant at Technical University of Munich, Computational Biology Lab, April 2017 – May 2018.
    Supervisor: Professor Julien Gagneur.
    – Developed multi-task ConvNets for predicting RNA-protein binding sites from RNA sequences. Keras
    – Developed a deep neural network module based on spline transformation to robustly model distances to various genomic landmarks which significantly increased state-of-the-art prediction accuracy of RNA-binding protein binding sites for 114 out of 123 proteins.
    – Published in bioinformatics Oxford academic journal: https://academic.oup.com/bioinformatics/advance-article/doi/10.1093/bioinformatics/btx727/4636216
    – Code: goo.gl/3yMY5w.

  • Research Intern at National Institute of Informatics, Prendinger Research Lab, Tokyo, Japan, September 2016 – March 2017.
    Supervisor: Professor Helmut Prendinger
    – Developed a novel aerial view dataset for the task of detecting concurrent actions of multiple humans; used “vatic” tool and “Amazon Mechanical Turk” for annotation. This resulted in a publication that will appear in Computer Vision and Pattern Recognition Workshops (CVPRW) proceedings. Dataset is now available at: http://okutama-action.org
    – Implemented new deep learning models for semantic segmentation task by adopting Residual Network into Fully Convolutional Networks. Knowledge Distillation technique was applied to further improve the performance. The final models were either faster by 50% or more accurate by 5%. This resulted in a publication in Elsevier Computer Vision and Image Understanding journal.
    – Implemented a deep learning model for action detection based on Single Shot MultiBox Detector (SSD) model.

  • Researcher at Applied Image and Signal Processing (AISP) research lab, University of Isfahan, September 2013 – 2015.
    Supervisor: Professor Amir Hassan Monadjem
    Developed and implemented a multi-scale/multi-directional Walsh-Hadamard transform for fast and robust texture feature extraction for image segmentation and texture analysis. Also involved in development of application which aimed to detect the drowsiness of a car driver’s eyes for android devices.

  • Isfahan Math House, Isfahan, Iran, 2008 – 2012
    – Researched on Game Theory. Designed a solution for solving combinatorial games by inverse rules. Also created a new combinatorial game. The Research outcome was Ranked 5th in the Kharazmi Youth Festival, biggest scientific competition among Iranian students.

Publications

  • Avsec, Z., Barekatain, M., Cheng, J., & Gagneur, J., “Modeling positional effects of regulatory sequences with spline transformations increases prediction accuracy of deep neural networks”. Bioinformatics, 2017.
  • Barekatain, M., Marti, M., Shih, H., Murray, S., Nakayama, K., Matsuo, Y., Prendinger, H., “Okutama-Action: An Aerial View Video Dataset for Concurrent Human Action Detection“. In Proc. Computer Vision and Pattern Recognition Workshops (CVPRW), Hawaii, USA, IEEE, 2017.
  • Holliday, A., Barekatain, M., Laurmaa, J., Kandaswamy, C., & Prendinger, H., “Speedup of Deep Learning Ensembles for Semantic Segmentation Using a Model Compression Technique”. Computer Vision and Image Understanding, Elsevier, 2017.
  • Mahdavinejada, M., Rezvan, M., Barekatain, M., Adibi, P., & Sheth, A., “Machine learning for Internet of Things data analysis: A survey”. Digital Communications and Networks, Elsevier, 2017.
  • Rezvan, M., Barekatain, M. Taghandiki K. Zaeri A.: “Applying an Innovative Semantic Sensor Network Model in Internet of Things”, 6th International Conference on ICT Convergence (ICTC 2015), Korea, IEEE, pages 324-328.
  • Rezvan, M., Barekatain, M.: “The Sensors Are Innovative in Internet of Things”, 8th International Conference, WICON 2014, Lisbon, Portugal, November 13-14, 2014, Revised Selected Papers, Springer LNICST 146 series, Pages 253-261

Honors and Awards

  • Won bronze medal by placing 3rd in Munich and 7th in total among 98 teams, at the German Collegiate Programming Contest (GCPC) 2016, Germany, 2016
  • Ranked 6th among all 60 teams and 1st among all teams of Munich in the ACM-ICPC FAU Wintercontest 2016, Germany, 2016
  • Ranked as the top student during all 8 semesters among students of Computer Engineering – Hardware, 2011 – 2015.
  • Ranked 29 in M.Sc. National University Entrance Exam for Iranian graduate studies among more than 300,000 applicants, Iran, spring 2015.
  • Received honorary admission from University of Isfahan graduate studies in Artificial Intelligence for outstanding academic success, Iran, spring 2015.
  • Ranked 29 in the 20th Iranian National Collegiate Scientific Olympiad in Computer Engineering, Iran, spring 2015.
  • Ranked 10th in the ACM-ICPC Asia Region Contest, Iran, fall 2014.
  • Ranked 9th in the ACM-ICPC Asia Region Contest, Iran, fall 2013.
  • Honorable Mention in the 2nd “Sharif University of Technology Open Robotics Competition” (Sharif Cup), Tehran, Iran, summer 2013.

Teaching Experiences

  • Teaching Assistant at Technical University of Munich, September 2018 – February 2019
    – Course: Introduction to Deep Learning – holding weekly tutorial classes for 200 Master students.
  • Teaching Assistant at Technical University of Munich, March 2017 – September 2017
    – Course: Statistical Modeling and Machine Learning – held weakly tutorial classes for 40 Master students.
  • Instructor at University of Isfahan, 2013-2014
    – Taught two summer courses on Algorithms and Data Structures for programming contests with 46 and 53 enrollees.

  • Teaching Assistant at University of Isfahan, 2012-2015
    – Courses: Artificial Intelligence, Algorithm Design, Computer Architecture, Advanced Programming I (two semesters), Data Structures, Computer Fundamentals (two semesters), Electrical Circuits.
    – Promoted to Head TA in 2015; defined final projects, led weekly meetings, and supervised other TAs.


Selected Projects

  • Face Detection and Facial Attribute Editing, 2018
    – Implemented a pipeline that first detects faces (using Single Shot MultiBox Detector) in an image and then changes specified facial attributes of each detected faces using a StarGAN, code link. TensorFlow, Pytorch
  • Anomaly Detection in Robot Time Series Data, 2016.
    – Implemented a semi-supervised model for detecting anomalies in robot time series data using Recurrent Neural Networks and Gaussian Mixture Models. Keras, Theano, Python
  • UEFA European Championship Predictor, 2016.
    Implemented an Ensemble of Classifiers model which predicts the outcomes of football matches given any information anterior to the match. Used SVM, Random Forest, AdaBoost and Logistic Regression classifiers. Python
  • Developed machine vision modules for BeagleBoard and Odroid-XU3 devices, 2015.
  • Solved Rectilinear Polygon Floor-planning problem using A* algorithm, fall 2014.
  • Implemented Mano Computer on FPGA Spartan III board using VHDL, spring 2014.
  • signalworks logo

    Signal Works: Designed a Direct Manipulation user interface for MATLAB, fall 2013.
    – Implemented a user interface which translates the input block diagram to MATLAB code for audio processing applications. JavaFX, MATLAB. Awarded “Honorable Mention” in the Research Week at the University of Isfahan.

  • Implemented Mano Computer with microprogrammed CPU using Logic Works, fall 2013.

Skills

  • Keras, TensorFlow, Python, Java, LATEX, UNIX/Linux, Git(Hub)
  • Basic: Pytorch, Theano, Caffe, R, C, C++, MATLAB, JavaFx

Selected Courses

Technical University of Munich:

  • Introduction to Deep Learning (2018)
  • Neuronal Deep Learning for Autonomous Systems (2017)
  • Machine Learning (2016)
  • Master Lab Course – Deep Learning for The Real World (2016)
  • Machine Learning for Computer Vision (2016)
  • Statistical Modeling and Machine Learning (2016)
  • Content Based Image Retrieval (2016)

University of Isfahan:

  • Numerical Analysis (2015): 20/20 [Top Mark]
  • Artificial Intelligence (2015): 18.25/20 [Top Mark]
  • Statistics and Probability (2013): 19.6/20 [Top Mark]
  • Calculus II (2012): 20/20 [Top Mark]
  • Computer Graphics (2015): 19.70/20 [Top Mark]
  • Advanced Programming II (2014): 20/20 [Top Mark]
  • Algorithm Design (2013): 20/20 [Top Mark]
  • Signals And Systems (2014): 19.70/20 [Top Mark]
  • Data Base (2015): 18.75/20 [Top Mark]
  • Data Structures (2013): 20/20 [Top Mark]
  • Discrete Structures (2012): 20/20 [Top Mark]

Extracurricular Activities

  • Member of Council; theinternetofthings.eu, 2014 – present.
  • Member of the organization committee of the first International Internet of Things event in Iran, September 2014.
    Description: The event was planned in two days; one day for a conference and one day for a workshop. Dr. Rob van Kranenburg, founder of European Internet of Things Council, was invited for this event.
  • Problem Designer for ACM-ICPC programming contests at University of Isfahan, Iran, 2013 – 2015.
  • Manager of ACM-ICPC Student Council Teaching Programs, 2013 – 2015.
  • Committee member of ACM-ICPC Student Council, University of Isfahan, 2013 – 2015.
  • Member of Isfahan Math House (IMH),  Isfahan, Iran, 2007 – 2011.
  • Bronze medalist in Isfahan Regional Chess Tournament, 2004.

Languages

  • English: Fluent
  • Persian/Farsi: Native
  • German: Basic

References

  • Available upon request.