KIT | KIT-Bibliothek | Impressum | Datenschutz

Topology-Matching Normalizing Flows for Out-of-Distribution Detection in Robot Learning

Feng, Jianxiang; Lee, Jongseok 1; Geisler, Simon; Günnemann, Stephan; Triebel, Rudolph 1
1 Institut für Anthropomatik und Robotik (IAR), Karlsruher Institut für Technologie (KIT)

Abstract:

To facilitate reliable deployments of autonomous robots in the real world, Out-of-Distribution (OOD) detection capabilities are often required. A powerful approach for OOD detection is based on density estimation with Normalizing Flows (NFs). However, we find that prior work with NFs attempts to match the complex target distribution topologically with naive base distributions leading to adverse implications. In this work, we circumvent this topological mismatch using an expressive class-conditional base distribution trained with an information-theoretic objective to match the required topology. The proposed method enjoys the merits of wide compatibility with existing learned models without any performance degradation and minimum computation overhead while enhancing OOD detection capabilities. We demonstrate superior results in density estimation and 2D object detection benchmarks in comparison with extensive baselines. Moreover, we showcase the applicability of the method with a real-robot deployment.


Volltext §
DOI: 10.5445/IR/1000168926
Veröffentlicht am 28.02.2024
Cover der Publikation
Zugehörige Institution(en) am KIT Institut für Anthropomatik und Robotik (IAR)
Publikationstyp Forschungsbericht/Preprint
Publikationsjahr 2023
Sprache Englisch
Identifikator KITopen-ID: 1000168926
Umfang 12 S.
Vorab online veröffentlicht am 11.11.2023
Schlagwörter Normalizing Flows, Out-of-Distribution, Robotic Introspection
Nachgewiesen in arXiv
Dimensions
Relationen in KITopen
KIT – Die Forschungsuniversität in der Helmholtz-Gemeinschaft
KITopen Landing Page