Thomas Altstidl — accepted AI-conference papers

2025

Stratify or Die: Rethinking Data Splits in Image Segmentation

Naga Venkata Sai Jitin Jami, Thomas Altstidl, Jonas Mueller, Jindong Li, Dario Zanca, Bjoern Eskofier +1

NeurIPS 2025poster

Random splitting of datasets in image segmentation often leads to unrepresentative test sets, resulting in biased evaluations and poor model generalization. While stratified sampling has proven effective for addressing label distribution imbalance in classification tasks, extending these ideas to se…

Cited by 0 Source ↗Scholar ↗

2024

On the Scalability of Certified Adversarial Robustness with Generated Data

Thomas Altstidl, David Dobre, Arthur Kosmala, Bjoern Eskofier, Gauthier Gidel, Leo Schwinn

NeurIPS 2024poster

Certified defenses against adversarial attacks offer formal guarantees on the robustness of a model, making them more reliable than empirical methods such as adversarial training, whose effectiveness is often later reduced by unseen attacks. Still, the limited certified robustness that is currently…

Cited by 0 Source ↗PDF ↗Scholar ↗

2023

$p$-value Adjustment for Monotonous, Unbiased, and Fast Clustering Comparison

Kai Klede, Thomas Altstidl, Dario Zanca, Bjoern Eskofier

NeurIPS 2023poster

Popular metrics for clustering comparison, like the Adjusted Rand Index and the Adjusted Mutual Information, are type II biased. The Standardized Mutual Information removes this bias but suffers from counterintuitive non-monotonicity and poor computational efficiency. We introduce the $p$-value adju…

Cited by 3 Source ↗PDF ↗code ↗Scholar ↗