Für eine korrekte Darstellung dieser Seite benötigen Sie einen XHTML-standardkonformen Browser, der die Darstellung von CSS-Dateien zulässt.

You are here


Dr. Jovita Lukasik

Room: 
H-A 7117/2
Mail: 
jovita.lukasik (at) uni-siegen.de
Address: 

Universität Siegen
Lehrstuhl für Computer Vision
Hölderlinstraße 3
57068 Siegen

I work as a PostDoc in the DFG research unit Learning to Sense.

My work on Neural Architecture Search during my PhD was  funded by the BMBF (Federal Ministry of Education and Research) in the project DeToL – Deep Topology Learning at the University of Mannheim.  I was also affiliated with MPII.

I was part of the organization team of the second NAS workshop @ ICLR 2021.
I am co-organizing a series of virtual seminars on AutoML.
I defended my PhD Thesis on "Topology Learning for Prediction, Generation, and Robustness in Neural Architecture Search" on July 13, 2023 at the University of Mannheim!

Publications

Improving Native CNN Robustness with Filter Frequency Regularization
Jovita Lukasik*, Paul Gavrikov*, Janis Keuper, Margret Keuper
Transactions on Machine Learning Research (TMLR), 2023
[pdf]
An Evaluation of Zero-Cost Proxies - from Neural Architecture Performance to Model Robustness
Jovita Lukasik, Michael Moeller, Margret Keuper
German Conference on Pattern Recognition (GCPR), 2023
[arXiv]
Differentiable Architecture Search: a One-Shot Method?
Jovita Lukasik*, Jonas Geiping*, Michael Moeller, Margret Keuper
AutoML Conference Workshops (AutoML), 2023
 
Neural Architecture Design and Robustness: A Dataset
Steffen Jung*, Jovita Lukasik*, Margret Keuper
International Conference on Learning Representations (ICLR), 2023
[website]
Learning Where to Look - Generative NAS is Surprisingly Efficient
Jovita Lukasik* , Steffen Jung*, Margret Keuper
European Conference on Computer Vision (ECCV), 2022
[arXiv]
Surrogate NAS Benchmarks: Going Beyond the Limited Search Spaces of Tabular NAS Benchmarks
Arber Zela*, Julien  Siems*, Lucas Zimmer*, Jovita Lukasik, Margret Keuper, Frank Hutter
International Conference on Learning Representations (ICLR), 2022
[arXiv]
Smooth Variational Graph Embeddings for Efficient Neural Architecture Search
Jovita Lukasik, David Friede, Aber Zela, Frank Hutter, Margret Keuper
International Joint Conference on Neural Networks (IJCNN), 2021
[arXiv]
Neural Architecture Performance Prediction Using Graph Neural Networks
Jovita Lukasik, David Friede, Heiner Stuckenschmidt, Margret Keuper
German Conference on Pattern Recognition (GCPR), 2020
[arXiv]