A two-stage sparse attention mechanism, selecting top regions then top pixels per query, improves accuracy on multiple medical imaging benchmarks with lower compute than full attention.
Advances in medical image analysis with vision transformers: A comprehensive review,
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
-
MedFormer: Hierarchical Medical Vision Transformer with Content-Aware Dual Sparse Selection Attention
A two-stage sparse attention mechanism, selecting top regions then top pixels per query, improves accuracy on multiple medical imaging benchmarks with lower compute than full attention.