Keuper Labs
← All publications

BioImage Computing Workshop at ECCV 26 · 2026

Unsupervised Visual Concept Bottlenecks for Interpretable Bioimage Classification

Linghui Liu, Henrike Stephani, Jördis Sieburg-Rockel, Stephanie Helmling, Andrea Olbrich, Janis Keuper

Concept Bottleneck ModelsExplainable AIBioimage AnalysisUnsupervised Learning

Why this publication matters

Biological image classifiers are easier to inspect when their decisions can be connected to visible intermediate concepts. This work investigates learning those concepts without manually defining every one. The aim is to make useful visual evidence part of the prediction process and support more understandable bioimage analysis.

Abstract

The original abstract will be added when the full paper is available.

Figures

Additional figures and captions will be added when the full paper is available.

Cite this paper

Download .bib 
@inproceedings{liu2026unsupervisedvisualconcept97,
  title = {{Unsupervised Visual Concept Bottlenecks for Interpretable Bioimage Classification}},
  author = {Linghui Liu and Henrike Stephani and Jördis Sieburg-Rockel and Stephanie Helmling and Andrea Olbrich and Janis Keuper},
  booktitle = {BioImage Computing Workshop at ECCV 26},
  year = {2026},
  url = {https://openreview.net/pdf?id=hois3HoeYg}
}

Figures and abstract are reproduced from the linked research sources. Credit remains with the authors and publishers.