ScSAM fuses frozen SAM and MAE embeddings with a fusion module and class prompt encoder, reporting state-of-the-art organelle segmentation on BetaSeg islet datasets.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
ScSAM: Debiasing Morphology and Distributional Variability in Subcellular Semantic Segmentation
ScSAM fuses frozen SAM and MAE embeddings with a fusion module and class prompt encoder, reporting state-of-the-art organelle segmentation on BetaSeg islet datasets.