Source Themes

(Predictable) Performance Bias in Unsupervised Anomaly Detection

We discovered that performance bias between demographic subgroups in unsupervised anomaly detection follows certain patterns, making it highly predictable.

Felix Meissen, Svenja Breuer, Moritz Knolle, Alena Buyx, Ruth Müller, Georgios Kaissis, Benedikt Wiestler, Daniel Rückert

Unsupervised Pathology Detection: A Deep Dive Into the State of the Art

So far the largest comparison of anomaly detection models with various paradigms on four medical imaging datasets.

Ioannis Lagogiannis, Felix Meissen, Georgios Kaissis, Daniel Rückert