I was born and raised in Kilkis – a small town of Central Macedonia, Greece – but have lived in several other places such as Ioannina, Athens, Xanthi, Heraklion, Providence, Zürich, Seattle, Boston and Washington DC. I'm a computing professional with a special interest and expertise in the World Wide Web, and I have a keen interest in sports, especially road running. I love traveling (who doesn't?) and I try to combine it with outdoor activities such as hiking, snowboarding, snorkeling and boating.
As a kid, I spent many years learning foreign languages (French, German, Italian) that I can hardly speak any more. Later, I studied computer science at the Universities of Ioannina and Crete, and then I kind of switched to cognitive science at Brown University. A few yeas ago I completed a flight school for single engine land & sea airplanes, and nowadays I'm trying to study a bit of business administration. I've always been intrigued by the process of learning, both academically and personally, and I'm a strong supporter of lifelong education.
[PhD thesis] Visual Recognition with a Large Scale Network of Dynamical Systems
[MSc thesis] A Network of Networks Approach to Computational Modeling of Pop-out Phenomena in Visual Attention
[MSc thesis] An Image Retrieval Platform Based on a Biologically Inspired Architecture
[BSc thesis] Human-Computer Interaction with Gesture Recognition
It started as a hobby while in grad school and ended up becoming a full fledged business. A combination of right timing and risk seeking made Greek-Movies the first search engine of - exclusively Greek - video streaming, and a site that was serving one million visitors per month. As a full stack developer I was taking care of everything, from system administration and back-end development to front-end responsive design and adertising. It may not be the trend any more but it gave me the privilege of experiencing software development to its fullest extent and I loved it. Working on every aspect of a website, from the engineering all the way to the marketing and the regulatory compliance, offers a unique perspective on the Internet.
I initially worked on neural networks because I found it an excellent approach to understanding (and implementing) intelligence. Then I did research on computer vision because I thought it offered the perfect testbed for experimenting with computational intelligence theories. Finally, I switched to cognitive science because I came to believe that deciphering intelligence requires the devoted study of the computational aspects of the human mind. I'm fascinated by large scale dynamical neural networks and I hope they will help us discover many of the answers we're looking for.
Ersatz Brain Group
Modeling of visual recognition based on large scale integration of dynamical neural networks. High performance computing with MPI and C++ on an IBM 166-node linux cluster. Watch a few visualizations of a system that models several features of the human visual recognition process. Each of these simulations models about 1 billion neural synapses and requires about 100 CPU cores for several minutes.
Laboratory for Engineering Man/Machine Systems – Computer Vision
Use of satellite imagery for the detection and evaluation of landscape and urban changes
3D Shape Perception Lab
Bayesian modeling of visual depth perception for inference under limited cognitive resources
Ersatz Brain Group
Modeling of popout phenomena in visual attention with a network of dynamical neural networks. Watch two visualizations of the network dynamics that model the Müller-Lyer illusion and the popout phenomena in visual attention
Computational Vision and Robotics Laboratory
Development of a content based image retrieval system based on a multi-agent architecture
Computational Vision and Robotics Laboratory
Development of a robotic soccer platform including the mechanical, electronic, and software components. Watch the first stages of a prototype robot that was developed from scratch
Test and evaluation of video conference platforms for distance learning
Development of web-based Neural Network simulators for a distance learning course. Try the following NN web simulators; the code is 20 years old but still working (if your browser allows Java applets)
Development of a gesture recognition system for the translation of the Greek Sign Language. It uses a 5DT data glove interface and is trained with a neural network.
Some projects that I'm currently working on in my free time. They're quite small but filling
A few projects that I voluntarily developed in the past (note: some of them have changed completely since then)