Omar Faig Orujlu

AI and Computer Vision Engineer

Combining expertise in electrical engineering and AI to drive innovation in computer vision and robotics

Software Developer in C#

Experienced in developing robust applications using C# and .NET framework.

Certificates

Azure Databricks & Spark for Data Engineers(PySpark / SQL) Generative AI with Diffusion Models by Nvidia AWS Certified Machine Learning Specialty 2024 - Hands On!

I am a recent graduate with a Master's degree in Robotics, Cognition, and Intelligence from Technische Universität München (TU Munich). I have a solid foundation in Electrical Engineering, having completed my B.Eng. from HTW Berlin. My Master’s thesis focused on domain adaptation for point cloud data using deep learning methods, highlighting my skills in computer vision and AI technologies. Additionally, my Bachelor’s thesis involved implementing an autoencoder with triplet loss, which further solidified my expertise in machine learning and neural networks. I am proficient in Python, PyTorch, TensorFlow, and have experience with Azure Databricks and other similar ML frameworks.

I am also an experienced Software Developer in C#, with a strong background in developing robust applications using the .NET framework. My projects have included building scalable software solutions, implementing complex algorithms, and optimizing performance for enterprise-level applications.

I am fluent in English and German, with both languages at the C1 level. I am eager to apply my knowledge and technical skills to contribute to innovative projects in the field of artificial intelligence, computer vision, and software development. With a passion for AI, computer vision, and software engineering, I am excited about the opportunities to drive technological advancements and solve complex problems in these dynamic fields.

12/2021 - Now

HARMAN Becker Automotive Systems GmbH

Working Student

Since December 2021, I have been working as a Working Student at HARMAN Becker Automotive Systems GmbH. During my time at the company, I have been involved in various projects within the automotive domain. I implemented a Python algorithm for reading texts from input images using OpenCV for preprocessing and Tesseract OCR for text extraction. Additionally, I developed a component management tool using ASP.NET Razor Pages, creating a webpage with HTML, CSS, and JavaScript, and maintained a database in SQLite.

Python OpenCV Tesseract OCR ASP.NET Razor Pages HTML CSS JavaScript C# SQLite

10/2020 - 02/2021

Robert Bosch GmbH

Internship

During my internship at Robert Bosch GmbH, I primarily focused on PLC programming. I developed programs in Structured Text (ST) using TwinCAT. My responsibilities included writing detailed documentation for the code to ensure clarity and maintainability. Additionally, I conducted both hardware and software tests to verify the functionality and reliability of the systems. This experience enhanced my skills in automation and provided me with valuable hands-on experience in an industrial setting.

PLC Programming Structured Text (ST) TwinCAT Documentation Hardware Testing Software Testing

Master's Thesis

3D Object Detection Optimization using Sim to Real Domain Adaptation with Neural Networks

My Master's thesis focused on researching domain adaptation and applying state-of-the-art point completion networks to learn the characteristics of real-world data and generate adapted datasets. This project highlights my expertise in 3D object detection and neural networks. The source code and details of the project are available on my github profile.

3D Object Detection Domain Adaptation Neural Networks

Practical Course

Overparameterized Sigmoid Autoencoder

In this practical course, I researched the memory behavior in overparameterized sigmoid autoencoders. The autoencoders were trained with different activation functions including sigmoid, tanh, ReLU, and Leaky ReLU. I added various degrees of noise to the input to observe the memory behavior of the network and analyzed the Jacobian matrices for further insights. This work utilized PyTorch, NumPy, scikit-learn, and other relevant frameworks and libraries.

Autoencoders PyTorch NumPy scikit-learn Neural Networks

Practical Course

Robust-Odometry-Estimation for-RGB-D Cameras

In this course, we extensively used C++ to develop robust odometry estimation techniques for RGB-D cameras. The project involved creating a 3D representation from RGB-D data and accurately estimating the camera's motion over time. This work required a deep understanding of both computer vision and robotics principles.

C++ RGB-D Odometry

Personal Project [In Progress]

Vision Analytics Hub

In this project I am developing a Vision Analytics Hub. The idea is to have a central hub for all vision related tasks. The hub will provide a simple and intuitive interface for users to upload data, train models, and analyze results. The hub will be developed using FastAPI and the users will be able to use the hub via a web interface.

Python FastAPI React YOLOv8

Personal Project [In Progress]

Weather App

In this project, I am developing a weather application in C# using the .NET framework with Avalonia as the UI framework. This application is a clone of the Microsoft Weather app and will display weather forecasts. Additionally, it will feature an AI-based forecast system that predicts weather based on historical data. The project aims to provide an intuitive and visually appealing interface while leveraging AI to enhance the accuracy of weather predictions.

C# .NET AvaloniaUI

Personal Project [In Progress]

SLAM CPP

In this project, I am compiling the source code of all the famous SLAM (Simultaneous Localization and Mapping) papers. The goal is to create a comprehensive repository that includes implementations of various SLAM algorithms, providing a valuable resource for researchers and developers working in the field of robotics and computer vision. This project involves understanding and replicating the methodologies of seminal SLAM papers, ensuring accurate and efficient implementations.

C++ SLAM Robotics