AGA builds token-to-patch and patch-to-token groups via a sparse similarity matrix with adaptive thresholds, and trains a medical image-text model with within-pair contrastive losses.
Multi-granularity cross-modal alignment for generalized medical visual representation learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
AGA: An adaptive group alignment framework for structured medical cross-modal representation learning
AGA builds token-to-patch and patch-to-token groups via a sparse similarity matrix with adaptive thresholds, and trains a medical image-text model with within-pair contrastive losses.