SegWithU treats uncertainty as perturbation energy via rank-1 probes in a post-hoc head for efficient single-pass risk-aware medical image segmentation, outperforming other single-forward-pass methods on ACDC, BraTS2024, and LiTS.
Deep deterministic uncer- tainty for semantic segmentation
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.CV 2verdicts
UNVERDICTED 2representative citing papers
Combines pre-trained features, Bayesian regression, and moment propagation to enable real-time epistemic uncertainty for semantic segmentation on embedded systems while preserving accuracy.
citing papers explorer
-
SegWithU: Uncertainty as Perturbation Energy for Single-Forward-Pass Risk-Aware Medical Image Segmentation
SegWithU treats uncertainty as perturbation energy via rank-1 probes in a post-hoc head for efficient single-pass risk-aware medical image segmentation, outperforming other single-forward-pass methods on ACDC, BraTS2024, and LiTS.
-
Uncertainty in Real-Time Semantic Segmentation on Embedded Systems
Combines pre-trained features, Bayesian regression, and moment propagation to enable real-time epistemic uncertainty for semantic segmentation on embedded systems while preserving accuracy.