I designed, developed and brought to market Radio Belgium, a native app to turn any iOS or Android device instantly into an amazing radio player.
Since launch, the app has been downloaded over 500 000 times with more than 150 000 monthly active users and frequently tops the App Store charts. The app was also featured on popular press sites, such as De Standaard and Humo.
Besides being an interesting project to sharpen my technical skills, it also allowed me to gain hands-on experience in a variety of non-technical fields such as product management, business development and marketing
Inspired by the succes of Angry Birds and Cut The Rope, I developed and released my own physics-based puzzle game during high school.
Tapaway was my first foray into iOS development and made use of the Cocos2d game engine, which I integrated with the Box2D physic engine. This combination allowed me too learn the ropes of programming and memory management in Objective-C(++).
The end result was an enjoyable game with over 30 handcrafted levels and received great reviews from numerous players. Since this was mainly a learning project, I stopped supporting it when iOS transitioned to 64-bit. A small demo of the game is still available on YouTube.
Watch demo on YouTube
After completing courses on artificial intelligence and machine learning, I was excited about the huge potential of these technologies. I was especially intrigued by deep learning and thus decided to do my master’s thesis in this domain.
Specifically, the title of my master’s thesis is automatic comic colorization and restoration through deep learning. Briefly worded, the goal is to fully automatically colorize degraded black-and-white pages from old comics. To do so, I’m building upon the powerful pix2pix architecture that proposes conditional adversarial networks as a general-purpose solution to image-to-image translation problems.
Completing this thesis, allowed me to immerse myself into recent deep learning research. I also gained a lot of hands-on experience with applying state-of-the-art deep learning techniques to solve challenging problems.
Writing a path tracer is a great way to become more acquainted with a language, since it touches upon so many different common aspects of a language. Besides, it is very cool to watch your own program making a scene come alive.
I’m still working on some key features and plan on publishing the full source code on GitHub in the near future.