News

• "Optical System Characterization in Fourier Ptychographic Microscopy" by J. Meshreki, S. M. Kazim, and I. Ihrke has been accepted in Optics Continuum, Optica, 2024. 

• “How do training methods influence the utilization of vision models?” by P. Gavrikov, S. Agnihotri, M. Keuper, and J. Keuper, “has been published in NeurIPS workshop on Interpretable AI: Past, Present and Future, 2024.

• "Improving stability during upsampling – on the importance of spatial context" by S. Agnihotri, J. Grabinski, and M. Keuper has been published in European Conference on Computer Vision, Springer, 2024. 

• "Learning the essential in less than 2k additional weights - a simple approach to improve image classification stability under corruptions" by K. Bäuerle, P. Müller, S. M. Kazim, I. Ihrke, and M. Keuper has been published in Transactions on Machine Learning Research, 2024.

• “Leveraging vignetting in fourier ptychographic microscopy” by J. Meshreki, S. M. Kazim, J. P. Schneider, M. Moeller, and I. Ihrke has been published in IEEE International Conference on Computational Photography (ICCP) - Posters, 2024.

• "CosPGD: An efficient white-box adversarial attack for pixel-wise prediction tasks" by S. Agnihotri, S. Jung, and M. Keuper has been published in Proceedings of the 41st International Conference on Machine Learning, R. Salakhutdinov, Z. Kolter, K. Heller, et al., Eds., ser. Proceedings of Machine Learning Research, vol. 235, PMLR, 21–27 Jul 2024, pp. 416–451.
  Video: https://icml.cc/virtual/2024/poster/34678
  Github: https://github.com/shashankskagnihotri/cospgd

• "Adaptive high dynamic range in fourier ptychographic microscopy" by J. Meshreki and I. Ihrke has been published in Journées Imagerie Optique Non-Conventionelle (JIONC), 2024.

• "Implicit representations for constrained image segmentation" by J. P. Schneider, M. Fatima, J. Lukasik, A. Kolb, M. Keuper, and M. Moeller has been published in International Conference on Machine Learning, 2024.
  Github:  https://github.com/jp-schneider/awesome

• "Surprisingly strong performance prediction with neural graph features" by G. Kadlecová, J. Lukasik, M. Pilát, et al. has been published in International Conference on Machine Learning, 2024.

• "Task driven sensor layouts – joint optimization of pixel layout and network parameters" by H. Sommerhoff, S. Agnihotri, M. Saleh, et al., has been published in International Conference on Computational Photography, 2024.

• "Variable layout cmos pixels for end-to-end learning in task specific image sensors" by B. Choubey, H. Sommerhoff, M. Moeller, and A. Kolb, has been published in International Conference on AI Circuits and Systems (AICAS), 2024

• "Modelling vignetting in fourier ptychographic microscopy" by J. Meshreki, J. P. Schneider, O. Martorell, M. Moeller, and I. Ihrke has been published in International Symposium on Computational Sensing, 2023.
  Video: https://www.youtube.com/watch?v=KgB8wSDMceU

• "Defect detection in bidirectional glass fabric reinforced thermoplastics based on 3-d-thz imaging" by A. Souliman, M. Kahl, D. Stock, M. Möller, B. Engel, and P. H. Bolívar has been published in IEEE Transactions on Terahertz Science and Technology, vol. 13, no. 3, pp. 209–220, 2023. doi: 10.1109/TTHZ.2023. 3247609.

• "Improving Native CNN Robustness with Filter Frequency Regularization" by Jovita Lukasik, Paul Gavrikov, Janis Keuper and Margret Keuper has been accepted and published in Transactions on Machine Learning Research 2023.

• "An Evaluation of Zero-Cost Proxies -- from Neural Architecture Performance to Model Robustness" by Jovita Lukasik, Michael Moeller and Margret Keuper has been accepted at the German Conference on Pattern Recongnition (GCPR).

• "On the unreasonable vulnerability of transformers for image restoration--and an easy fix" by Shashank Agnihotri, Kanchana Vaishnavi Gandikota, Julia Grabinski, Paramanand Chandramouli and Margret Keuper will be presented at the 4th Workshop on Adversarial Robustness In the Real World (AROW) ICCV 2023, Paris

• Our spokesperson, Michael Moeller, received a prestigeous Lamarr-Fellowship (https://www.land.nrw/pressemitteilung/renommierte-experten-fuer-kuenstliche-intelligenz-ausgezeichnet)

• "An Evaluation of Zero-Cost Proxies -- from Neural Architecture Performance to Model Robustness" by Jovita Lukasik, Michael Moeller and Margret Keuper will be presented at the German Conference on Pattern Recongnition.

• Our Mercator Fellow Felix Heide is receiving the ACM SIGGRAPH New Researcher Award!

• Our Mercator Fellow Wolfgang Heidrich is receiving the ACM SIGGRAPH Computer Graphics Achievement Award!

• John Meshreki, Jan Philipp Schneider, Onofre Martorell, Michael Moeller, and Ivo Ihrke, have their extended abstract "Modelling Vignetting in Fourier Ptychographic Microscopy" accepted at the International Symposium on Computational Sensing 2023.

• A. Souliman, M. Kahl, M. Möller, B. Engel, P. Haring Bolívar have their abstract "Subsurface Defect Detection And Classification In 3D THz Images Of Glass Fiber Reinforced Thermoplastic Based On 3D Convolutional Neural Network" accepted at the 48th IRMMW-THz 2023 Conference in Montreal, Canada