Cosmin Ionuț Bercea is a Researcher at the Albarqouni Lab led by Shadi Albarqouni at Helmholtz AI and the Technical University of Munich (TUM). His aim is to develop innovative deep Federated Learning algorithms for anomaly detection in medical images in a robust and privacy-preserved fashion.

He was granted a study scholarship by the e.V. Villigst and received his B.Sc. and M.Sc. in Computer Science from the Friedrich-Alexander University of Erlangen-Nuremberg (FAU), Germany, in 2015, and 2018, respectively. During his master, Bercea did an internship abroad at the Computer Vision Centre (CVC) in Barcelona, Spain, where he developed machine learning algorithms for action recognition. He completed his master thesis at Siemens Healthineers, where he worked with Olivier Pauly, Florin C. Ghesu and Prof. Andreas K. Maier on developing novel machine learning algorithms using shared memory for medical imaging. Since then, Bercea has worked as a research engineer at Bosch Corporate Research where he developed perception and behaviour prediction machine learning algorithms for self-driving cars. His current research interests include Precision Medicine, Anomaly Detection, Image Understanding and Federated Learning.


  • Deep Learning for Medical Image Analysis
  • Anomaly Detection
  • Federated Learning
  • Image Understanding


  • M.Sc. in Computer Science, 2018

    Friedrich-Alexander Universität Erlangen-Nürnberg (FAU), Germany

  • (Exchange Semester) in Computer Science, 2016

    Universitat Autonoma de Barcelona (UAB), Spain

  • B.Sc. in Computer Science, 2015

    Friedrich-Alexander Universität Erlangen-Nürnberg (FAU), Germany