Abdarhman Sayed

Abdarhman Sayed Mahmoud

Software Engineer | Natural Language Processing, Computer Vision and Deep Learning

Research Engineer at COM-IoT Technologies.

Heliopolis, Cairo, Egypt
(+2) 0100-357-1139
Arabic, English

Background


About

About

I'm a software engineer with extreme passion for research, Now I'm a research engineer in COM-IoT Technologies where I work on distracted driver detection and tracking, previously I was a part of the research team in Myocardial Solutions, Inc., my professional experience is mainly in research areas related to the usage of computer vision in 3D reconstruction of underwater scenes and cardiac segmentation from MRI, I'm also doing research in the use of NLP in enhancing text retrieval and query-focused multi-document summarization, and I'm a huge fan of Manchester United.
Research Area

Research Area

  • DeepSENC, Real-Time Left Ventricle Segmentation From Strain-Encoded MRI Using U-NET

    Oct 2015 - May, 2016 8 months

    We developed a novel pipeline for SENC images segmentation using a histogram based region proposal and a UNET for segmentation, and we were able to run this pipeline in real-time on a 4-core CPU.

    • Paper [ Under Review ]: [PDF]
  • Artificial Intelligence and Parallel Computing, GPU Accelerated Timetabling System, Gradution Project

    Oct 2015 - May, 2016 8 months

    An enhanced genetic algorithm-based timetabling system with incremental changes

    • Paper [ IEEE ICCES 2016 ]: [PDF]
    • Presentation (Including web architecture): [PDF]
  • Computer Vision and Machine Learning, Automatic Arabic Sign Language Recognition, Undergraduate research

    April 2015 - July, 2015 2 months

    The use of hidden markov model [HMM] in automatic recognition of sign language using preconstructed human body

Work Experience

Work Experience

  • Research Engineer Myocardial Solutions, Inc.

    Jan, 2018 - Present2 months

    • Researching deep learning techniques for the automatic segmentation of the myocardium from MRI .
    • Used encoder-decoder networks like SegNet, U-Net.
  • R&D SW Engineer | Computer Vision & Deep Learning MCS Enterprise

    Aug, 2016 - Jan 20176 months

    • Implemented a preprocessing tool for underwater image restoration to enhance feateure extraction, this tool is written in C++ .
    • Parallelized the implementation using CUDA and OpenMP.
    • Used a convolutional neural network (CNN) for better scene matching, achieved 86%, this technique had reduced searching time for matching scenes dramatically vs SIFT based techniques.
  • R&D SW Developer (Intern) , Valeo

    April, 2016 - Aug, 20165 Months

    • Built a tool that analyzes the code architecture for safety measurements before being deployed to the car’s hardware.
    • Used Understand's python API for analysis.
  • Undergraduate Internships

    • ASP.Net Backend Intern, NasHope
    • Universal Application Developer, Microsoft Egypt
    • Mobile Communications Intern, Vodafone Egypt
    • Softwre Testing Intern, Hewlett-Packard Egypt
Education

Education

  • Computer Systems and Engineering, Bachelor, Ain Shams University

    2011 - 2016

Awards

MOOCs

  • DEV288x Natural Language Processing (NLP) | Microsoft

    On: Feb, 2018

  • Improving Deep Neural Networks, Hyperparameter tuning, Regularization and Optimization | deeplearning.ai

    On: Dec 12, 2017

  • Neural Networks and Deep Learning | deeplearning.ai

    On: Oct 10, 2017

  • MIT 6.S094: Deep Learning for Self-Driving Cars | MIT

  • Text Retrieval and Search Engines | University of Illinois at Urbana-Champaign

    On: Apr 28, 2017

  • Introduction to Data Science in Python | University of Michigan

    On: Aug 15, 2017

  • CS224n: Natural Language Processing with Deep Learning | Stanfod University

  • Deep Learning 101 | IBM Cognitive Class

    On: May 12, 2017

  • CS231n: Convolutional Neural Networks for Visual Recognition | Stanfod University

  • Using Python to Access Web Data | University of Michigan

    On: Apr 08, 2012

Interests

Interests

  • Sports

    Football ( Manchester United )Swimming