Skip to content

Oszkár Józsa
Software Engineer, Robotics and Computer Vision Professional
I am a Robotics Engineer at Ascent Robotics, focusing on self-driving technology, machine learning and software development. Ascent is a quickly growing AI startup that aims to be the best in the field in Japan. The current focus is on AI research and simulation and machine learning platform development (Atlas Engine) with the long-term aim of a general purpose AI that will be used for drones, cars, spacecraft and anything that moves.
I am also a master’s student at Georgia Tech, participating in the OMSCS program (in partnership with Udacity). I am pursuing the Computational Perception & Robotics specialization and I plan to finish the program at the end of 2017.
Previous Positions
I have worked at 3 companies before my current position
2011-2017
  • Self-Driving Car Engineer at ZMP Inc. I was doing both software engineering ans system engineering, working on ZMP’s self driving technology called RoboCar (which is the testbed vehicle of the RobotTaxi). I mainly worked on Lidar technology (Velodyne, Konica Minolta, IBEO), both on system integration/calibration and software development including object detection and tracking algorithm research. I was also a self-driving car operator, testing the RoboCar and its systems on the streets of Tokyo.
  • Software Engineer at the Cognex Corporation, I was member of the Vision Products unit (Core Engineering team). I worked on front-end development of industrial computer vision software (Cognex VisionView, Cognex Designer) both on PC and embedded systems at the Budapest office of the Company. I was part of a large multi-site, multi-language, multi-product project spanning over 2.5 years from initial software design through final release.
  • Research Intern/Contractor at the Distributed Events Analysis Research Laboratory of the Hungarian Academy of Sciences Institute for Computer Science and Control. I mostly worked on urban 3D point cloud algorithms research and development, namely on the i4D project. Using point clouds collected with a Velodyne lidar, I worked on scene understanding, object detection and segmentation and 3D surface reconstruction. I also worked on 2D image processing research.

(see more detail on the Work Experience page)

Professional Interests
  • Software engineering
  • Computer vision and 2D/3D image processing
  • Robotics and autonomous vehicles
  • Space exploration
Professional Experience
I worked on several different projects and have broad knowledge of various technologies
I am quite agnostic in environment and languages and I like to learn new tools
  • Software Engineering and Development in various languages
  • Designing and implementing complex software systems, including distributed systems and real-time systems
  • Development on the Microsocft stack (Windows desktop & CE, Visual Studio, C#, TypeScript, XAML)
  • I’m also comfortable on Linux (I use Mint and Ubuntu)
  • C++ development, including 3rd party libs (Boost, CUDA, Qt, Point Clouds Library, OpenCV, CGAL, libSVM)
  • Various scripting languages (Matlab, Python, JavaScript, TypeScript, R)
  • Web application development (HTML5, JavaScript), and a few JS frameworks (AngularJS, RequireJS, jQuery, node.js)
  • 3D point clouds and 2D image processing
  • SVM machine learning methods
  • Multispectal imagery
Examples of Earlier Work