Neural Architecture Performance Prediction Using Graph Neural Networks
Published in DAGM German Conference on Pattern Recognition, 2020
Jovita Lukasik, David Friede, Heiner Stuckenschmidt, Margret Keuper
Download here
Posts by Collection
open_positions
portfolio
publications
Smooth Variational Graph Embeddings for Efficient Neural Architecture Search
Published in International Joint Conference on Neural Networks, 2021
Jovita Lukasik, David Friede, Arber Zela, Frank Hutter, Margret Keuper
Download here
DARTS for Inverse Problems: a Study on Stability
Published in NeurIPS 2021 Deep Inverse Workshop, 2021
Jonas Geiping, Jovita Lukasik, Margret Keuper, Michael Moeller
Download here
Surrogate NAS Benchmarks: Going Beyond the Limited Search Spaces of Tabular NAS Benchmarks
Published in The Tenth International Conference on Learning Representations, 2022
Arber Zela, Julien Niklas Siems, Lucas Zimmer, Jovita Lukasik, Margret Keuper, Frank Hutter
Download here
Learning Where to Look – Generative NAS Is Surprisingly Efficient
Published in European Conference on Computer Vision, 2022
Jovita Lukasik, Steffen Jung, Margret Keuper
Download here
Neural Architecture Design and Robustness: A Dataset
Published in The Eleventh International Conference on Learning Representations, 2023
Steffen Jung, Jovita Lukasik, Margret Keuper
Download here
Differentiable Architecture Search: a One-Shot Method?
Published in AutoML Conference 2023 Workshops, 2023
Jovita Lukasik, Jonas Geiping, Michael Moeller, Margret Keuper
Download here
An Evaluation of Zero-Cost Proxies - from Neural Architecture Performance to Model Robustness (oral)
Published in DAGM German Conference on Pattern Recognition, 2023
Jovita Lukasik,Michael Moeller, Margret Keuper
Download here
Implicit Representations for Image Segmentation
Published in Unireps@NeurIPS 2023, 2023
Jan Philipp Schneider, Mishal Fatima, Jovita Lukasik, Andreas Kolb, Margret Keuper, Michael Moeller
Download here
Improving Native CNN Robustness with Filter Frequency Regularization
Published in Transactions on Machine Learning Research, 2023
Jovita Lukasik, Paul Gavrikov, Janis Keuper, Margret Keuper
Download here
Implicit Representations for Constrained Image Segmentation
Published in Proceedings of the 41st International Conference on Machine Learning, 2024
Jan Philipp Schneider, Mishal Fatima, Jovita Lukasik, Andreas Kolb, Margret Keuper, Michael Moeller
Download here
Surprisingly Strong Performance Prediction with Neural Graph Features
Published in Proceedings of the 41st International Conference on Machine Learning, 2024
Gabriela Kadlecová, Jovita Lukasik, Martin Pilát, Petra Vidnerová, Mahmoud Safari, Roman Neruda, Frank Hutter
Download here
An Evaluation of Zero-Cost Proxies - from Neural Architecture Performance to Model Robustness
Published in International Journal of Computer Vision, 2024
Jovita Lukasik,Michael Moeller, Margret Keuper
Download here
Can we talk models into seeing the world differently?
Published in The Thirteenth International Conference on Learning Representations (accepted), 2025
Paul Gavrikov, Jovita Lukasik, Steffen Jung, Robert Geirhos, Muhammad Jehanzeb Mirza, Margret Keuper, Janis Keuper
Download here
talks
Inference given Summary Statistics
Published:
Invited Talk by Prof. Dr. Janis Keuper
Topology Learning for Prediction, Generation, and Robustness in Neural Architecture Search
Published:
Invited Talk by Prof. Dr. Eddy Ilg
Topology Learning for Prediction, Generation, and Robustness in Neural Architecture Search
Published:
Invited Talk by “Learning 2 Sense” as lecture series
AutoML, Trustworthiness and Alignment
Published:
Invited as Panelist by Prof. Dr. Frank Hutter
An Evaluation of Zero-Cost Proxies - from Neural Architecture Performance to Model Robustness
Published:
Presentation of “An Evaluation of Zero-Cost Proxies - from Neural Architecture Performance to Model Robustness” at the Conference
Multi-Objective Performance Prediction for Neural Architectuer Search
Published:
Presentation of “Multi-Objective Performance Prediction for Neural Architectuer Search” at the Conference
Neural Architecture Search - Towards Prediction beyong (CIFAR-10 Accuracy) Benchmarks
Published:
Invited Talk by Prof. Dr. Carola-Bibiane Schoenlieb
Bridging the Gap between Industry and Academia
Published:
Invited as Panelist by Prof. Dr. Frank Hutter