About me

I am currently a Software Engineer at Mismatch which creates software for automated live streaming and tactical analysis of basketball games powered by computer vision and machine learning. My work involves:

  • Implementation of computer vision modules, e.g. camera calibration, OCR, background subtraction
  • Porting research code into production-ready machines (code optimization, GPU utilization, communication via REST APIs)

I have obtained a Diploma (Integrated Master) in Electrical and Computer Engineering from the University of Patras, Greece. My expertise is in Computer Software and Programming. Following my undergraduate studies, I enrolled for the Postgraduate program "Data-Driven Computing and Decision Making" at the Department of Computer Engineering and Informatics (CEID) where I successfully completed my postgraduate studies and obtained a Master of Science degree. The postgraduate courses made me built a solid mathematical foundation of machine learning and the projects of the courses helped me improve my research skills in machine learning and data science. During my postgraduate studies, I have supported as a teaching assistant the following undergraduate courses:

  • Intelligent Systems and Robotics (CEID_NE552) (March 2021 - June 2021)
  • Databases (CEID_ΝΥ334) - (November 2019 - February 2021)
  • Object-Oriented Programming (CEID_NY134) - (March 2020 - June 2020)

In my master thesis, I had the chance to be supervised by Dr. Konstantinos Chatzilygeroudis who introduced me to the fascinating world of robotics and intelligent systems. Particularly, I performed research on Sim2real Methods for Visual-based Imitation Learning. My master thesis raised some important research questions which I am willing to explore deeper by pursuing a PhD in the aforementioned field. My main research interests include Computer Vision, Robot Learning and Machine Learning.

Aside from academic and industrial work, I have a diploma in Byzantine Music and I serve as a cantor in my neighborhood's church. My interests include cycling, digital photography, observational astronomy and reading popular-science books. Some of the books that I really enjoyed reading and I recommend are:

  • The Ethical Algorithm: The Science of Socially Aware Algorithm Design by Aaron Roth & Michael Kearns (2019) (Also check the authors' presentation here)
  • Human Compatible: AI and the Problem of Control by Stuart Russell (2019)

Education

Ph.D. candidate in Sim2Real Methods for Visual-based Imitation Learning
March 2022 - Now

Department of Computer Engineering and Informatics (CEID), Department of Mathematics
University of Patras, Greece
Master of Science (M.Sc.) in Data-Driven Computing and Decision Making
October 2019 - September 2021

Department of Computer Engineering and Informatics (CEID), Department of Mathematics
University of Patras, Greece
Grade: 9.46 / 10
Integrated Master (300 ECTS) in Electrical and Computer Engineering
September 2012 - April 2019

Department of Electrical and Computer Engineering, University of Patras, Greece
Grade: 7.79 / 10
Diploma Thesis: Registration of satellite images
Supervisor: Prof. Athanassios Skodras

Technical Skills

AI skills

This chart represents my scores on seven essential AI skills based on the Workera assessment tests. These standardized tests, developed by a team of AI experts led by Dr. Andrew Ng and Kian Katanforoosh from deeplearning.ai, evaluate the ability to carry out tasks involved in AI projects including data engineering, modeling, deployment, business analysis and AI infrastructure. For more information about the tests you can check here.


Website Building Skills

I have spent my free time attending online courses and reading tutorials about Front-End Web Development (HTML, CSS, JavaScript, JQuery) and I have been awarded a certificate for completing successfully "Responsive Web Design" course series provided by freeCodeCamp.org.

In addition, I have designed and I currently manage the following websites using WordPress:

Projects

Registration of satellite images

This project was mandatory for my graduation from the Department of Electrical and Computer Enginnering, University of Patras, Greece. It was supervised by Prof. Athanassios Skodras and carried out at Digital Signal and Image Processing Laboratory. You can view the poster of this diploma thesis here and a version of the diploma thesis in Greek here.

Abstract

Image registration is a process in which images that exhibit geometric deviations are mapped to each other in a common coordinate system. In the present work, the usefulness of the registration to satellite imaging is examined. For this purpose, a feature-based method was implemented to calculate the geometric transformation that aligns the images. The method was implemented with the OpenCV and FLANN libraries. In particular, the performance of SIFT and SURF detectors and brute-force and multiple randomized K-D trees matching methods were evaluated. Finally, two satellite imaging applications were implemented in which image registration is applied, change detection on the surface of Earth and image mosaicking for mapping a bigger area.