Dr. Jens Behrmann

Research Areas

• Deep Learning
• Inverse problems
• Imaging mass spectrometry

Projects

1. AGENS - Analytical-generative Networks for System Identification (01.04.2020 - 31.03.2023)
2. Studie zur Qualitätsbewertung, Standardisierung und Reproduzierbarkeit von Daten der bildgebenden MALDI-Massenspektrometrie – MALDISTAR (01.07.2019 - 30.06.2022)
3. Neural networks in MALDI imaging (since 01.10.2016)
4. BMBF-MaDiPath: Mass spectrometric profiling and grading for oncologic routine applications in pathology (01.10.2015 - 30.09.2018)
5. Entwicklung eines Digital-Staining-Verfahrens als pathologisch-histologisches Diagnosewerkzeug auf Basis der MALDI-Imaging-Technologie (01.07.2014 - 30.06.2016)

Courses (Selection)complete list

1. Oberseminar: Deep Learning, Inverse Probleme und Datenanalyse (Wintersemester 2020/2021)
2. Oberseminar: Deep Learning, Inverse Probleme und Datenanalyse (Sommersemester 2020)
3. Mathematical Foundations of Machine Learning (Sommersemester 2019)
4. Seminar zu Neuronalen Netzen (Wintersemester 2018/2019)
5. Seminar zu Neuronalen Netzen (Wintersemester 2016/2017)

1. Übungen Computerpraktikum (Wintersemester 2016/2017)
2. Übungen Computerpraktikum (Wintersemester 2015/2016)

Theses (Selection)complete list

1. Statistical Inversion of the Radon Transform with Normalizing Flows (Jakob Lehmann)
2. Residuale Transformationen und Divergenzen in Bayes’schen Neuronalen Netzen (Niklas Koenen)
3. Long-Term Time Series Forecasting and Uncertainty Estimation with Bayesian Neural Networks (David Erzmann)
4. Application of Neural Networks For Solving Inverse Problems (Alexander Denker)
5. Ein universelles Approximationstheorem für einschichtige neuronale Netze (Meira Iske)

Publications (Selection)complete list

1. C. Janßen, T. Boskamp, L. Hauberg-Lotte, J. Behrmann, S. Deininger, M. Kriegsmann, K. Kriegsmann, G. Steinbuß, H. Winter, T. Muley, R. Casadonte, J. Kriegsmann, P. Maaß.
   Robust subtyping of non-small cell lung cancer whole sections through MALDI mass spectrometry imaging.
   Proteomics - Clinical Applications, PRCA2208 , 2022.

   DOI: 10.1002/prca.202100068

2. J. Behrmann, M. Schmidt, J. Wildner, P. Maaß, S. Schmale.
   Purity Assessment of Pellets Using Deep Learning.
   German Success Stories in Industrial Mathematics, H. Bock, K. Küfer, P. Maaß, A. Milde, V. Schulz (Eds.), Mathematics in Industry, pp. 29-34, Springer Verlag, 2022.

   DOI: 10.1007/978-3-030-81455-7_6

3. A. Denker, M. Schmidt, J. Leuschner, P. Maaß, J. Behrmann.
   Conditional Normalizing Flows for Low-Dose Computed Tomography Image Reconstruction.
   ICML Workshop on Invertible Neural Networks, Normalizing Flows, and Explicit Likelihood Models, 18.07-18.07.2020, Vienna, Austria.
   

   online at: https://invertibleworkshop.github.io/accepted_papers/index.html

4. F. Tramer, J. Behrmann, N. Carlini, N. Papernot, J. Jacobsen.
   Fundamental Tradeoffs between Invariance and Sensitivity to Adversarial Perturbations.
   International Conference on Machine Learning (ICML), 12.07 - 18.07.2020, Vienna, Austria.
   

   online at: https://arxiv.org/abs/2002.04599

5. J. Behrmann, P. Vicol, K. Wang, R. Grosse, J. Jacobsen.
   Understanding and Mitigating Exploding Inverses in Invertible Neural Networks.
   Zur Veröffentlichung eingereicht.

   online at: https://arxiv.org/abs/2006.09347