UncertainSAM: Fast and Efficient Uncertainty Quantification of the Segment Anything Model

Timo Kaiser, Thomas Norrenbrock, Bodo Rosenhahn
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:28670-28688, 2025.

Abstract

The introduction of the Segment Anything Model (SAM) has paved the way for numerous semantic segmentation applications. For several tasks, quantifying the uncertainty of SAM is of particular interest. However, the ambiguous nature of the class-agnostic foundation model SAM challenges current uncertainty quantification (UQ) approaches. This paper presents a theoretically motivated uncertainty quantification model based on a Bayesian entropy formulation jointly respecting aleatoric, epistemic, and the newly introduced task uncertainty. We use this formulation to train USAM, a lightweight post-hoc UQ method. Our model traces the root of uncertainty back to under-parameterised models, insufficient prompts or image ambiguities. Our proposed deterministic USAM demonstrates superior predictive capabilities on the SA-V, MOSE, ADE20k, DAVIS, and COCO datasets, offering a computationally cheap and easy-to-use UQ alternative that can support user-prompting, enhance semi-supervised pipelines, or balance the tradeoff between accuracy and cost efficiency.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-kaiser25a, title = {{U}ncertain{SAM}: Fast and Efficient Uncertainty Quantification of the Segment Anything Model}, author = {Kaiser, Timo and Norrenbrock, Thomas and Rosenhahn, Bodo}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {28670--28688}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/kaiser25a/kaiser25a.pdf}, url = {https://proceedings.mlr.press/v267/kaiser25a.html}, abstract = {The introduction of the Segment Anything Model (SAM) has paved the way for numerous semantic segmentation applications. For several tasks, quantifying the uncertainty of SAM is of particular interest. However, the ambiguous nature of the class-agnostic foundation model SAM challenges current uncertainty quantification (UQ) approaches. This paper presents a theoretically motivated uncertainty quantification model based on a Bayesian entropy formulation jointly respecting aleatoric, epistemic, and the newly introduced task uncertainty. We use this formulation to train USAM, a lightweight post-hoc UQ method. Our model traces the root of uncertainty back to under-parameterised models, insufficient prompts or image ambiguities. Our proposed deterministic USAM demonstrates superior predictive capabilities on the SA-V, MOSE, ADE20k, DAVIS, and COCO datasets, offering a computationally cheap and easy-to-use UQ alternative that can support user-prompting, enhance semi-supervised pipelines, or balance the tradeoff between accuracy and cost efficiency.} }
Endnote
%0 Conference Paper %T UncertainSAM: Fast and Efficient Uncertainty Quantification of the Segment Anything Model %A Timo Kaiser %A Thomas Norrenbrock %A Bodo Rosenhahn %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-kaiser25a %I PMLR %P 28670--28688 %U https://proceedings.mlr.press/v267/kaiser25a.html %V 267 %X The introduction of the Segment Anything Model (SAM) has paved the way for numerous semantic segmentation applications. For several tasks, quantifying the uncertainty of SAM is of particular interest. However, the ambiguous nature of the class-agnostic foundation model SAM challenges current uncertainty quantification (UQ) approaches. This paper presents a theoretically motivated uncertainty quantification model based on a Bayesian entropy formulation jointly respecting aleatoric, epistemic, and the newly introduced task uncertainty. We use this formulation to train USAM, a lightweight post-hoc UQ method. Our model traces the root of uncertainty back to under-parameterised models, insufficient prompts or image ambiguities. Our proposed deterministic USAM demonstrates superior predictive capabilities on the SA-V, MOSE, ADE20k, DAVIS, and COCO datasets, offering a computationally cheap and easy-to-use UQ alternative that can support user-prompting, enhance semi-supervised pipelines, or balance the tradeoff between accuracy and cost efficiency.
APA
Kaiser, T., Norrenbrock, T. & Rosenhahn, B.. (2025). UncertainSAM: Fast and Efficient Uncertainty Quantification of the Segment Anything Model. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:28670-28688 Available from https://proceedings.mlr.press/v267/kaiser25a.html.

Related Material