2019
License plate recognition · Candid ITS
Development work on an adaptive number plate recognition engine and camera-based recognition systems.
The person behind Antal.AI
I’m Antal.
I like making computer vision work.
An independent engineering workshop, built on more than a decade of image processing, research and product development.
Curiosity is where
the projects begin.
I’m a software engineer in Hungary, specializing in computer vision, augmented reality and machine learning. My professional work has included embedded camera systems, automotive navigation, image processing and face analysis.
Antal.AI is where I develop and share my independent projects alongside that work. A question becomes an experiment, then a working engine, a demo and documentation that someone else can build on.
I value that freedom to explore. I also care about making the result understandable and useful to people with real engineering problems.
Why I work this way
A small workshop.
A clear handover.
I build reusable engineering foundations and make them available as self-serve downloads. Full source code, documentation and browser demos help you evaluate the work and take responsibility for your own integration.
This keeps the business sustainable for me and the expectations clear for you. I can keep developing, and you can inspect and adapt what you buy.
- See it working. Try the available demos with your own camera.
- Understand the details. Read the public code references before buying an engine.
- Know the boundaries. Review what is included, what requires your own development, and the license terms.
- Build independently. Integration support, custom development and ongoing maintenance are not included.
Selected professional background
Experience behind the experiments.
These are examples from my professional and research background. They give context to my work; they are separate from the downloadable Antal.AI products.
2018–2019
Gaze-controlled navigation · NNG
Led a team to win an internal innovation competition with a gaze-controlled point-of-interest search concept, then presented the solution to customers at CES 2019 in Las Vegas.
2017–2019
Augmented reality & hybrid navigation
Worked on augmented reality for head-up displays and on systems combining mobile and embedded navigation.
2014
Image processing across platforms
Developed an image processing library for mobile platforms and worked on sports performance analysis and intelligent surveillance.
2013
Audience measurement & mapping
Worked on camera-based audience measurement, road condition mapping, underwater mapping and a microcontroller-based navigation robot.
2012
Embedded vision & early innovation
Worked on real-time visitor counting and depth-camera sign language recognition. Won the “Jump! Be Innovative!” competition launched by TEVA and the University of Debrecen with a mobile application concept.
Watch the augmented reality demonstration ↗
A practical way to get in touch.
The demos, FAQ and product documentation answer the most common evaluation questions.
For an essential question that is not covered, email antal@antal.ai with the engine name, your intended platform and the specific point you need clarified. Replies are not guaranteed; social messages are not a support channel.
Have a look at what I’m building.
The demos and source packages are the best introduction.