Semantic segmentation models produce label flips within foreground regions under correlation shift, quantified by a new Flip diagnostic and an entropy-based flip-risk score.
ResNet-50 exhibits non-trivial FG-Miss across regimes (and an increase under𝜌=0.95), alongside measurable FG-Flip
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Right Regions, Wrong Labels: Semantic Label Flips in Segmentation under Correlation Shift
Semantic segmentation models produce label flips within foreground regions under correlation shift, quantified by a new Flip diagnostic and an entropy-based flip-risk score.