Metraipolis › archive › dc_8a8840f09524
SpatialUQ: Post-Hoc Uncertainty Quantification from Spatial Consistency in Black-Box Vision Models — arXiv:2610.09498
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.09498
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T11:11:06.314Z
- SHA-256 of the archived text
e620708de711f240ec4af96430a798048811bae53aef98e32f8443d6375d856e- Stable URL of this page
https://metraipolis.com/documento/dc_8a8840f09524
How to cite
SpatialUQ: Post-Hoc Uncertainty Quantification from Spatial Consistency in Black-Box Vision Models — arXiv:2610.09498 — arXiv (en), https://arxiv.org/abs/2610.09498 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_8a8840f09524 (Metraipolis archive dc_8a8840f09524, sha256 e620708de711f240ec4af96430a798048811bae53aef98e32f8443d6375d856e)
BibTeX
@misc{dc8a8840f09524,
title = {SpatialUQ: Post-Hoc Uncertainty Quantification from Spatial Consistency in Black-Box Vision Models — arXiv:2610.09498},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.09498},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_8a8840f09524; sha256 e620708de711f240ec4af96430a798048811bae53aef98e32f8443d6375d856e},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.09498. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Clinical vision models are often deployed as frozen black boxes with no access to internals, retraining, or ground truth at inference time. We introduce \textbf{SpatialUQ}, a post-hoc uncertainty method using only output probabilities. It measures the Jensen-Shannon divergence between the global prediction and the mean of five fixed spatial crops in six deterministic forward passes. The premise is simple, trustworthy predictions are spatially consistent. On NIH ChestX-ray14 (DenseNet-121, $N{=}25{,}596$), our Multicrop Uncertainty Score (MUS) reaches $0.784$ failure-detection AUC versus $0.664$ for MC-Dropout ($p{<}10^{-6}$) at one-fifth the compute, with native calibration ($\text{SCE}{=}0.049$ vs.\ $0.127$ for $\ell_1$), the best-calibrated among methods above 0.78 AUC. A supervised fusion of MUS with entropy, confidence, and $\ell_1$ reaches $0.832$, outperforming a five-member ensemble ($0.813$). MUS scales with model quality, reaching $0.899$ with BiomedCLIP ($\rho = 0.846$), while this relationship remains meaningful in-distribution ($\rho = 0.523$) but breaks down under severe distribution shift (VinBigData, $\rho = 0.027$). MUS is well-suited to diffuse findings but is less dependable for small focal lesions such as nodules. Code and experimental materials are publicly available at this https URL .
Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.09498 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-08T10:14:06.094Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).