DiracGAN
An interactive web application based on Which Training Methods for GANs do actually Converge?
My name is Johannes Heeg and I am a doctoral student in mathematics at the Technische Universität Ilmenau. I am supervided by Karl Worthmann and focus on topics around active learning and the Koopman operator framework.
In the past, I studied Computational Engineering and Mathematics at the Technical University of Darmstadt and the University of Stuttgart and hold a master’s degree in Computational Engineering. During my studies, I focused on robot learning and wrote my master’s thesis on ”Task-space exploration in robot reinforcement learning” under the supervision of Jan Peters. Prior to joining the TU Ilmenau, I was a research intern at Davide Scaramuzza’s Lab at the University of Zurich, where I worked on differentiable simulation for learning quadrotor control.
In my free time, I like to sing and play music, and I am a member of several ensembles. You can check out a barbershop performance or ask me about the next concert dates if you are interested.
Learning Quadrotor Control From Visual Features Using Differentiable Simulation, 2024
Johannes Heeg, Yunlong Song, Davide Scaramuzza
Task Space Exploration in Robot Reinforcement Learning, 2023
Master's Thesis supervised by Prof. Jan Peters, Ph.D. at Intelligent Autonomous Systems Group, TU Darmstadt
Modeling of Thermochemical Manifolds With Machine Learning Methods, 2020
Bachelor's Thesis supervised by Prof. Dr.-Ing. Christian Hasse at Simulation of Reactive Thermo-Fluid Systems, TU Darmstadt