Award-winning Poster Presentation on Machine Learning by TUD Physics Students

Johann Voigt has won the prize for the best poster presentation at the online conference “Offshell 2021”. The topic of his poster was “Artificial Neural Networks on FPGAs for Real-Time Energy Reconstruction of the ATLAS LAr Calorimeters“.

His results on the implementation of artificial neural networks on programmable electronic circuits, so-called FPGAs, were part of his Master thesis at the Institute of Nuclear and Particle Physics. By applying machine learning methods, the energy measurement of the ATLAS particle detector at the Large Hadron Collider (LHC) at CERN shall be improved. The artificial neural networks detect signals of photons and electrons which are absorbed in the Liquid-Argon calorimeters of the ATLAS experiment. Together with a team of scientists from Dresden and Marseille, Johann Voigt could show that the neural networks operate on FPGAs as precisely as on regular computers. This technology shall be applied at the upgrade of the ATLAS experiment for the high-luminosity phase of the LHC starting in 2027.

In the meantime, Johann Voigt has started his PhD studies at the IKTP to continue his research on artificial intelligence on FPGAs.

Reference: ATL-LARG-SLIDE-2021-312