PROTON: Prototype-Based Test-Time Online OOD Detection for Medical VLMs
Medical vision-language models (VLMs) enable zero-shot clinical image classification, yet reliably detecting out-of-distribution (OOD) inputs at deployment remains an open problem. No static scoring method works across all shift types: Maximum Concept Matching (MCM) on FLAIR achieves 76.4% AUROC for far-OOD but only 42.4% for covariate shifts such as ultra-wide-field fundus images, effectively random. We trace this to a structural mismatch: covariate-shifted inputs are indistinguishable from in-distribution samples in softmax space, yet occupy distinct regions in the VLM embedding space. To exploit this untapped signal, we propose PROTON (PROtotype-based Test-time ONline OOD detection), a...
| Authors | Abhijit Das, Nichula Wasalathilaka, Yifan Lu et al. |
|---|---|
| Primary category | cs.CV |
| Published | 18 June 2026 |
| Last updated at source | 18 June 2026 |
Connected in the knowledge graph
- abhijit2k01/proton-benchmarksDataset · cited by
Models and datasets referencing this paper
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.