Jonathan (iony) Mikler

Image
Welcome to my personal website!
  • I am a Robotics Engineer in love with the field.
  • Currently pursuing my Masters in Autonomous Systems at the Technical University of Denmark.
  • The code to some of my projects can be found here.
  • My CV is available here: CV
  • LinkedIn: LinkedIn

    • This website is (always) a work in progress… This place serves a double purpose.

    1. As an archive of work I’ve done before, for now with the purpose of showing potential employers what I’ve done.
    2. As a place to share some of the things I’ve learned and found interesting.

    Professional Background

    Please check my CV for a more detailed description of my professional background. Here are some highlights:

    1. Robotics Engineer Student @ Agri-Robot Currently
    2. Robotics Engineer @ Blue White Robotics 2020-2022
    3. Test Engineer @ Blue White Robotics 2019-2020

    Academic Background

    I’ve taken some interesting courses so far:

    1. Advanced Deep Learning for Computer Vision

      Main project: Detection Transformer for Monocular 3D Object Detection. Check the poster here. The code is available here and notes are here

      • Partially disconnecting the depth-estimator lead to better performance and lower variance during training.
      • Proposing use of disentangled losses to improve kearning (inspired by Cube R-CNN).
      • Replaced the visual encoder with a faster one from RT-DETR to explore real-time usage.
      • Reduced Backbone size to compare performance and speed.

      Other projects:

      1. Visual Transformer for Image Classification
      2. Diffusion Models to Generate (guided and not) 16x16 game sprites.
      3. Novel View Generation with NeRFs (my favourite)

    2. Artificial Intelligence for Multi-Agent Systems

      Main Project: Windowed Hierarchichal Cooperative A* Multi-Agent Planning.video, Code

      • Developed a multi-agent planner that uses a windowed hierarchical approach to solve cooperative pathfinding problems.
      • Uses a hierarchical approach to solve the problem, where a high-level planner assigns subgoals to agents and an agent-level planner solves the pathfinding problem.
      • To solve conflicts, the planner uses a windowed approach, where agents are allowed to replan their paths within a time frame if they collide with other agents.

    3. Advanced Image Analysis Learning

      Main Project: Probabilistic Deformable Models for Image Segmentation. Poster

    4. Perception for Autonomous Systems

    5. Logical Theories for Uncertainty and Planning

      Main Project: Epistemic Logic for Reinforcement Learning. Report

    6. On the topic of RL: I’m currently studying how Reinforcement Learning works in Multi-Agent Systems, on a special course dedicated to the subject. Nothing too crazy yet, but I’m sure I’ll have something to share (hopefully).
    1. On the topic of Logic: I studied how epistemic logic could be used to help in the learning process of reinforcement learning agents. This was the final work of that class. The work is available here and was done together with Elle McFarlane a great AI engineer if you need one.
    2. A Midsummer’s Bayesian Dream: Something cool I learned in a logical theories class (Under process…)

    Other projects:

    I develop on a need-to-have basis these two shell toolboxes:

    1. Generic Shell Toolbox: Shell Toolbox
    2. ROS Shell Toolbox: ROS Shell Toolbox