Skip to content

People

Prof. Dr. Martin Mundt (he/him)

Head of the OWL-ML group

Email: martin.mundt _at_ uni-bremen.de

Martin Mundt is the OWL-ML research group leader at the University of Bremen. He is also a board member of directors at the non-profit organization ContinualAI and core-organizer at Queer in AI.He currently serves as Diversity, Equity and Inclusion (DEI) chair at CoLLAs 2025, was previously DEI Chair at AAAI-24 and Review Process Chair for CoLLAs-24.

Prior to joining the University of Bremen, Martin was a junior research group leader and visiting professor at TU Darmstadt and hessian.AI. He holds a PhD in computer science and a Master’s of Physics from Goethe University Frankfurt.

Martin’s work focuses on lifelong machine learning to render AI systems more adaptive, inclusive, robust, and sustainable. His respective work has received several distinctions, including best paper at FAccT-23, student paper highlight at AISTATS-24, an issue cover at journal of imaging, a best computer science lecture and a PhD thesis award

We are currently hiring two fully-funded PhD students on various topics related to lifelong machine learning – Application deadline of December 10th 2024

PhD Students

Roshni Kamath (she/her)


Email: roshni.kamath
_at_ tu-darmstadt.de

Roshni Kamath is currently a PhD student in the OWL-ML Lab at the Computer Science Department of the TU Darmstadt & hessian.AI. Previously, she worked as a Research Associate and AI Consultant at Forschungs-zentrum Jülich. She worked as a Software Engineer before pursuing her Masters in Artificial Intelligence from Katholieke Universiteit (KU) Leuven.

Subarnaduti Paul
(he/him)


Email: subarnaduti.paul
_at_ tu-darmstadt.de

Subarnaduti Paul decided to follow his passion for research by joining the OWL-ML Lab as a Ph.D. student after successfully graduating from TU Munich in the summer of 2022. His field of expertise lies in Electronic Systems & AI. Previously, he was working as a Research Intern for the Innovation Lab at Bosch Rexroth. His Thesis was able to propose reliable solutions that can integrate deep learning into traditional industrial processes. He will bring his previous experiences to instill a new picture in the paradigm of Open World Learning.

Research Assistants

  • Lars-Joel Frey: Master thesis, “Analyzing and Unlearning Hate in Machine Learning Models”. Previously also Bachelor thesis, “Towards Unsupervised Federated Continual Machine Learning“, completed in June 2022

Former members

  • Qi Li: Research assistant in 2023-2024 for project “CAIMAN: Continual Artificial Intelligence Models for Atmospheric Convection in Subtropical Regions“
  • Robin Menzenbach: Master thesis, “Self-Expanding Variational Autoencoders“, completed in March 2024. Former teaching assistant in winter 2022/23
  • Nick Lemke: Master thesis, “Distribution-Aware Replay for Continual MRI Segmentation“, co-supervised with Anirban Mukhopadhyay – GRIS TU Darmstadt, completed in March 2024
  • Patrick Vimr: Master thesis, “Continual Causal Knowledge Distillation“, completed in December 2023
  • Aleksandar Tatalovic: Bachelor thesis, “Continual Reinforcement Learning by Merging Models“, completed in October 2023
  • Tobias Gockel: Master thesis, “Knowledge Distillation for Continual Learning in Sum Product Networks“, completed in September 2023
  • Pranav Sureshkumar: Summer intern 2023 – DAAD WISE scholarship
  • Uranik Berisha: Master thesis, “Progressive Probabilistic Circuits“, completed in May 2023
  • Laura Boyette: Master thesis, “Deep Continual Learning with Intentional Forgetting“, completed in May 2023
  • Zhanke Liu: Master thesis, “Self-supervised Learning for Financial Time-Series Prediction via Transformers“, completed in April 2023
  • Dhruvin Vadgama: teaching assistant for continual machine learning course tutorials, winter semester 2022/23
  • Yves Neyraud: Master thesis, “Active Latent Space Packing for Variational Open World Learning“, completed in February 2022
  • Jesse-Jermaine Richter: Master thesis, “Continual Learning with Dataset Distillation“, completed in October 2022 (co-supervised with Kristian Kersting)
  • Sebastian Seer: Master thesis, “Learning Neural Network Latent Space Distributions with Probabilistic Circuits to Address the Overconfidence Challenge” (co-supervised with Kristian Kersting), completed in July 2022