Vahagn Yeghikyan, Developer in Yerevan, Armenia

Vahagn Yeghikyan

Computer Vision Developer

Location
Yerevan, Armenia
Toptal Member Since
November 16, 2021

Vahagn is a software engineer with 10+ years of experience and expertise in C++ and Python. His primary focus has been computer vision, but he has also worked on .NET services in C# and web-based crawlers with Python and Java. Vahagn has a strong mathematical background and holds a PhD in theoretical physics. He loves challenging projects and thrives in focused and collaborative environments.

Vahagn is available for hire
Hire Vahagn

Portfolio

SONY
C++, Standard Template Library (STL), IP Networks, Algorithms, Data Structures
NVision
C++11, OpenCV, TensorFlow, SQL, Azure, Azure SQL, Amazon Web Services (AWS)...
IntelinAir
OpenCV, TensorFlow, Python 3, Algorithms, Ubuntu Linux, Linux, C++...

Experience

Python - 8 yearsC#.NET - 7 yearsOpenCV - 5 yearsComputer Vision - 5 yearsC++ - 5 yearsTensorFlow - 5 yearsMachine Learning - 5 yearsSDKs - 3 years

Location

Yerevan, Armenia

Availability

Full-time

Preferred Environment

Linux, Ubuntu Linux, CLion, PyCharm

The most amazing...

...thing I've designed is a content-based image indexing and retrieval application with classic computer vision, machine learning, clouds, Docker, and Kubernetes.

Work Experience

2023 - 2023

Senior R&D Engineer

SONY
  • Developed an algorithm for an adaptive jitter buffer.
  • Tracked and fixed bugs using GitLab as a reporting tool.
  • Merged the algorithm into the master branch, which is now part of a large application.
Technologies: C++, Standard Template Library (STL), IP Networks, Algorithms, Data Structures
2018 - 2023

R&D Engineer

NVision
  • Designed and developed a scalable image indexing and retrieval application.
  • Researched state-of-the-art approaches for image indexing and proposed new solutions for the known problems.
  • Communicated with end users and designed and proposed solutions for the integration of our system with their platform.
  • Trained a variation of HiDDeN network as discussed in an academic paper (https://arxiv.org/abs/1807.09937).
Technologies: C++11, OpenCV, TensorFlow, SQL, Azure, Azure SQL, Amazon Web Services (AWS), Azure Blob Storage API, Azure Table Storage, Amazon S3 (AWS S3), AWS SDK, Google Cloud, Docker, Python 3, C#, Algorithms, Ubuntu Linux, Linux, C++, SDKs, Computer Vision, Google Mock (GMock), Python, Image Processing, C#.NET, Machine Learning, HTML, JavaScript, Django, Angular, APIs, Scraping, Data Scraping, Web Scraping
2017 - 2018

R&D Engineer

IntelinAir
  • Trained a cloud shadow detection network on aerial imagery.
  • Proposed a solution for an atmospheric effect compensation tool for infrared aerial imagery.
  • Completed multiple small tasks related to the IntelinAir platform.
Technologies: OpenCV, TensorFlow, Python 3, Algorithms, Ubuntu Linux, Linux, C++, Computer Vision, Python, Machine Learning, APIs
2015 - 2017

Senior Full-stack Developer

Gile
  • Developed a scalable application for large-scale web crawling and data mining.
  • Created Python and C# services for large-scale data manipulation.
  • Built MapReduce applications for producing reports from the crawled data.
Technologies: Python, Django, C#.NET, Java, Hadoop, HBase, Python 3, C#, Web Crawlers, Large-scale Web Crawlers, HTML, JavaScript, Angular, MapReduce, APIs, Scraping, Data Scraping, Web Scraping
2014 - 2015

Senior Full-stack Developer

BetConstruct
  • Developed an online scratching game that implemented the coin-flipping by telephone algorithm.
  • Tracked and fixed bugs using Jira as a reporting tool.
  • Completed minor tasks on BetConstruct's database ORM written in C++.
Technologies: C#.NET, ASP.NET MVC, SQL, MySQL, C++, C#, HTML, JavaScript, Jira, Algorithms, APIs

Experience

Idem

https://idem.vision/
Powered by machine learning technology, Idem is an intelligent tool that rapidly scans the web and finds uses of your original copyrighted images and videos, even matching doctored or edited versions.

GagoUI

https://github.com/vahagnIV/gago-ui
A small visual tool written in C++ to easily calibrate a stereo camera rig. It requires installing LibGago, a separate library (also written by me) that provides functionality for reading from and synchronizing two cameras. The existing tools provided with OpenCV and MATLAB are command-line and require some manual work.

Reimplementation of ORB SLAM3

https://github.com/vahagnIV/orb_slam3
A library for simultaneous localization and mapping (SLAM) for autonomous systems such as robots. This is a reimplementation of ORB SLAM3 library by Carlos Campos, Richard Elvira, Juan J. Gómez Rodríguez, José M. M. Montiel, and Juan D. Tardos.

Skills

Languages

Python 3, C++, C#, Python, C#.NET, Java, SQL, HTML, JavaScript

Libraries/APIs

OpenCV, TensorFlow, Standard Template Library (STL)

Other

Matrix Algebra, Physics, Scraping, Data Scraping, Web Scraping, Algebra, Linear Algebra, Abstract Algebra, Geometry, Advanced Physics, Computer Vision, Algorithms, Image Processing, Machine Learning, Web Crawlers, Large-scale Web Crawlers, APIs, SDKs, Simultaneous Localization & Mapping (SLAM), Robot Operating System (ROS), IP Networks, Data Structures

Frameworks

Google Mock (GMock), Qt, Django, Hadoop, ASP.NET MVC, Angular

Tools

AWS SDK, Jira

Platforms

Linux, Ubuntu Linux, Azure

Paradigms

MapReduce

Storage

HBase, MySQL

Education

2008 - 2011

PhD in Physics

Yerevan State University - Yerevan, Armenia