Tassilo Wald

About me

I recently completed my Ph.D. at the German Cancer Research Center (DKFZ), advised by Klaus H. Maier-Hein where I worked with many great colleagues, including Fabian Isensee and Paul Jäger. My research focuses on representation learning and foundation models, spanning self-supervised learning and vision-language models. I’m interested in understanding what deep networks learn and in building representations that transfer across tasks, with 3D medical imaging as the primary domain.

During my internship at Microsoft Research in Cambridge, UK, I worked with Fernando Pérez-García on 3D vision-language models for radiology report generation. The resulting model improved on the prior state of the art by +20 Macro-F1 and is now used by Microsoft’s collaborators at Mayo Clinic. At DKFZ, I led pre-training efforts for The Human Radiome Project and created or maintain open-source tools used across the field, including nnU-Net, OpenMind, and nnssl.

News

• [Sep. 2026] New blog post: Going from Primus to Primus V3, on what makes ViTs hard in 3D medical imaging and how we got Primus level with the best CNNs.
• [Sep. 2026] Our nnFoundation preprint on 3D foundation models for radiology is now online!
• [Sep. 2026] I gave a talk at the "Foundation Models for General CT Image Diagnosis" webinar.
• [Jul. 2026] I defended my PhD at the German Cancer Research Center (DKFZ)!
• [Jul. 2026] COLIPRI was accepted to ECCV 2026!
• [Apr. 2026] Primus, a ViT for 3D medical images, was accepted to TMLR!

Publications [Google Scholar]

1. Preprint
   nnFoundation: 3D Foundation Models for Radiology

   Constantin Ulrich Harsy^*, Tassilo Wald^*, Karol Gotkowski^*, Yannick Kirchhoff^*, Marcel Knopp^*, Maximilian Rokuss^*, …, Paul F. Jäger, Philipp Vollmuth, Fabian Isensee, Klaus H. Maier-Hein (82 authors)

   arXiv preprint, 2026.

   PDF Project Page BibTeX The Human Radiome Project
2. ECCV 26
   Comprehensive language-image pre-training for 3D medical image understanding

   Tassilo Wald, Ibrahim Ethem Hamamci, Yuan Gao, Sam Bond-Taylor, Harshita Sharma, Maximilian Ilse, Cynthia Lo, Olesya Melnichenko, Anton Schwaighofer, Noel C. F. Codella, Maria Teodora Wetscherek, Klaus H. Maier-Hein, Panagiotis Korfiatis, Valentina Salvatelli, Javier Alvarez-Valle, Fernando Pérez-García

   European Conference on Computer Vision 2026

   PDF Code Project Page BibTeX Microsoft Internship Project
3. TMLR 26
   Primus: Enforcing Attention Usage for 3D Medical Image Segmentation

   Tassilo Wald^*, Saikat Roy^*, Fabian Isensee^*, Constantin Ulrich, Sebastian Ziegler, Dasha Trofimova, Raphael Stock, Michael Baumgartner, Klaus Maier-Hein

   Transactions on Machine Learning Research, 05/2026

   PDF Code Project Page BibTeX
4. ICCV 25
   An OpenMind for 3D medical vision self-supervised learning

   Tassilo Wald^*, Constantin Ulrich^*, Jonathan Suprijiadi^*, Sebastian Ziegler, Michal Nohel, Robin Peretzke, Klaus Maier-Hein

   International Conference on Computer Vision

   PDF Code Project Page BibTeX
5. ER-x
   Enhancing deep learning methods for brain metastasis detection through cross-technique annotations on SPACE MRI

   Tassilo Wald^*, Benjamin Hamm^*, Julius C Holzschuh, Rami El Shafie, Andreas Kudak, Balint Kovacs, Irada Pflüger, Bastian von Nettelbladt, Constantin Ulrich, Michael Anton Baumgartner, Philipp Vollmuth, Jürgen Debus, Klaus H Maier-Hein, Thomas Welzel

   European Radiology Experimental

   PDF Project Page BibTeX
6. Preprint
   RadioActive: 3D Radiological Interactive Segmentation Benchmark

   Tassilo Wald^*, Constantin Ulrich^*, Emily Tempus^*, Maximilian Rouven Rokuss, Paul F. Jäger, Klaus Maier-Hein

   Preprint

   PDF Project Page BibTeX
7. CVPR 25
   Revisiting MAE pre-training for 3D medical image segmentation

   Tassilo Wald^*, Constantin Ulrich^*, Stanislav Lukyanenko, Andrei Goncharov, Alberto Paderno, Leander Maerkisch, Paul F. Jäger, Klaus Maier-Hein

   The IEEE/CVF Conference on Computer Vision and Pattern Recognition 2025

   PDF Project Page BibTeX Highlight
8. ICLR 25
   Resi: A comprehensive benchmark for representational similarity measures

   Tassilo Wald^*, Max Klabunde^*, Tobias Schumacher^*, Klaus H. Maier-Hein, Markus Strohmaier, Florian Lemmerich

   Thirteenth International Conference on Learning Representations

   PDF Code Project Page BibTeX
9. NeurIPS 24
   Decoupling Semantic Similarity from Spatial Alignment for Neural Networks

   Tassilo Wald, Gregor Köhler, David Zimmerer, Stefan Denner, Michael Baumgartner, Fabian Isensee, Priyank Jaini, Klaus H. Maier-Hein

   Neural Information Processing Systems - (NeurIPS), 2024.

   PDF Code Project Page BibTeX
10. MICCAI 24
    nnu-net revisited: A call for rigorous validation in 3d medical image segmentation

    Tassilo Wald^*, Fabian Isensee^*, Constantin Ulrich^*, Michael Baumgartner^*, Saikat Roy, Klaus H. Maier-Hein, Paul Jäger

    Medical Image Computing and Computer-Assisted Intervention - (MICCAI), 2024.

    PDF Code Project Page BibTeX
11. SCIS 23
    Exploring new ways: Enforcing representational dissimilarity to learn new features and reduce error consistency

    Tassilo Wald, Constantin Ulrich, Fabian Isensee, David Zimmerer, Gregor Koehler, Michael Baumgartner, Klaus H. Maier-Hein

    Published at the ICML 2023 Workshop on Spurious Correlations, Invariance, and Stability

    PDF Project Page BibTeX
12. MIDL short 23
    SAM.MD: Zero-shot medical image segmentation capabilities of the Segment Anything Model

    Saikat Roy^*, Tassilo Wald^*, Gregor Köhler^*, Maximilian R. Rokuss^*, Nico Disch^*, Julius Holzschuh^*, David Zimmerer^*, Klaus H. Maier-Hein

    Medical Imaging with Deep Learning (MIDL), short paper track, 2023.

    PDF Project Page BibTeX
13. BVM 23
    Extending nnU-Net is all you need

    Fabian Isensee^*, Constantin Ulrich^*, Tassilo Wald^*, Maier-Hein Klaus

    Bildverarbeitung für die Medizin (BVM), 2023.

    PDF Project Page BibTeX Technical Report of our 1st place challenge participation in AMOS2022.
14. NOA 22
    Automated detection and quantification of brain metastases on clinical MRI data using artificial neural networks

    Irada Pflueger^*, Tassilo Wald^*, Fabian Isensee, Marianne Schell, Hagen Meredig, Kai Schlamp, Denise Bernhardt, Gianluca Brugnara, Claus Peter Heußel, Juergen Debus, Wolfgang Wick, Martin Bendszus, Klaus H. Maier-Hein, Philip Vollmuth

    Neuro-Oncology Advances, 4(1), 2022

    PDF Code BibTeX

Publications (Co-authored)

1. WACV 24
   RecycleNet: Latent Feature Recycling Leads to Iterative Decision Refinement.

   Gregor Köhler, Tassilo Wald, Constantin Ulrich, David Zimmerer, Paul F. Jaeger, Jörg K. H. Franke, Simon Kohl, Fabian Isensee, Klaus H. Maier-Hein

   IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2024.

   PDF Project Page BibTeX Early accept
2. MICCAI 23
   MultiTalent: A Multi-dataset Approach to Medical Image Segmentation.

   Constantin Ulrich, Fabian Isensee, Tassilo Wald, Maximilian Zenk, Michael Baumgartner, Klaus H. Maier-Hein

   Medical Image Computing and Computer Assisted Intervention - (MICCAI), 2023.

   PDF Code Project Page BibTeX Oral presentation
3. MICCAI 23
   cOOpD: Reformulating COPD classification on chest CT scans as anomaly detection using contrastive representations

   Silvia D. Almeida^*, Carsten T. Lüth^*, Tobias Norajitra, Tassilo Wald, Marco Nolden, Paul F. Jaeger, Claus P. Heussel, Jürgen Biederer, Oliver Weinheimer, Klaus Maier-Hein

   Medical Image Computing and Computer Assisted Intervention - (MICCAI), 2023.

   PDF Code Project Page BibTeX
4. BVM 23
   Taming Detection Transformers for Medical Object Detection

   Marc K. Ickler^*, Michael Baumgartner^*, Saikat Roy, Tassilo Wald, Klaus H. Maier-Hein

   Bildverarbeitung für die Medizin (BVM), 2023.

   PDF Project Page BibTeX Nominated for Best Paper Award & Oral
5. IV 21
   Temporal Feature Networks for CNN based Object Detection

   Michael Weber, Tassilo Wald, J. Marius Zollner

   IEEE Intelligent Vehicles Symposium, 2021

   PDF Project Page BibTeX

Challenges & Awards

1. Award
   Award: Extraordinary Contributions Award, CVPR 2025 Interactive 3D Biomedical Image Segmentation Challenge (ALLDATA track)

   Fabian Isensee, Maximilian Rokuss, Lars Krämer, Stefan Dinkelacker, Ashis Ravindran, Florian Stritzke, Benjamin Hamm, Tassilo Wald, Moritz Langenberg, Constantin Ulrich, Jonathan Deissler, Ralf Floca, Klaus Maier-Hein

   Hosted by: CVPR 2025 Workshop on Foundation Models for Interactive 3D Biomedical Image Segmentation.

   Challenge Code Certificate Technical Report for nnInteractive
2. 2nd place
   Challenge: Head and Neck Tumor Segmentation for MR-Guided Applications (HNTS-MRG) 2024, Task 2 (mid-RT segmentation)

   Jessica Kächele, Maximilian Zenk, Maximilian Rokuss, Constantin Ulrich, Tassilo Wald, Klaus Maier-Hein

   Participating teams: 15

   Hosted by: Medical Image Computing and Computer Assisted Intervention - (MICCAI), 2024.

   Challenge Leaderboard
3. 1st place
   Challenge: Mediastinal Lymph Node Quantification (LNQ): Segmentation of Heterogeneous CT Data

   Jonathan Deissler^*, Tassilo Wald^*

   Participating teams: 13

   Hosted by: Medical Image Computing and Computer Assisted Intervention - (MICCAI), 2023.

   Challenge Leaderboard Certificate
4. Top performer
   Challenge: Tumor Detection, Segmentation and Classification on Automated 3D Breast Ultrasound (TDSC-ABUS) 2023

   Raphael Stock, Michael Baumgartner, Yannick Kirchhoff, Maximilian R. Rokuss, Jonathan Deissler, Nico Disch, Julius Holzschuh, Gregor Koehler, Saikat Roy, Tassilo Wald, David Zimmerer, Klaus Maier-Hein

   Hosted by: Medical Image Computing and Computer Assisted Intervention - (MICCAI), 2023.

   Challenge Certificate Overall top performer
5. 1st place
   Challenge: Abdominal Multi-Organ Segmentation (AMOS) Challenge 2022

   Fabian Isensee^*, Constantin Ulrich^*, Tassilo Wald^*

   Participating teams: 20

   Hosted by: Medical Image Computing and Computer Assisted Intervention - (MICCAI), 2022.

   Challenge Leaderboard Technical Report
6. 1st place
   Hackathon: AI-HERO Hackathon for Energy-Efficient AI

   Tassilo Wald^*, Michael Baumgartner^*, Gregor Köhler^*

   Participating teams: 4

   Hosted by: Helmholtz Information & Data Science Academy (HiDA), 2022.

   Challenge

Miscellaneous

• Organizer of the SSL3D Challenge (Self-Supervised Learning for 3D Medical Imaging) at MICCAI 2025.
• Invited talk at Cohere on self-supervised learning for 3D medical imaging, in a series that has also featured speakers such as Lucas Beyer.
• Organizer of heidelberg.ai since 2023, a local AI research community with more than 3,000 members.
• Outstanding reviewer at CVPR 2025.
• Regular reviewer for NeurIPS, ICLR, ICML, CVPR and ICCV/ECCV.
• Maintainer of nnU-Net (8,800+ GitHub stars) since 2023, the standard framework for 3D dense prediction.
• Author of nnssl, a reference implementation of self-supervised learning methods for 3D vision.
• Reviewer for the UniReps workshop at NeurIPS 2023.
• Speaker at the advanced deep learning tutorial at BVM 2022, on multi-task learning for medical image analysis.

Contact

Email: tassilo.wald[at]gmail[dot]com

Powered by Jekyll and based off Minimal Light theme.