WRSWorld Research StateContribute
WORLD RESEARCH STATE

Swin Transformer: hierarchical shifted-window attention

Contribution grade

Vision architectures · Human-designed research method or resource

Reference score180,000Provisional editorial estimate of this exact contribution; its grade is assessed separately.

What changed

Used hierarchical patch merging and alternating shifted local-attention windows to connect regions while limiting per-layer spatial attention cost.

Scientific scope

Used hierarchical patch merging and alternating shifted local-attention windows to connect regions while limiting per-layer spatial attention cost. Inherits transformers, local attention and multiscale vision. Linear complexity refers to a fixed-window design as spatial size scales, not arbitrary global attention or every end-to-end operation. Benchmarks are source-specific.

Evidence

The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Inherits transformers, local attention and multiscale vision. Linear complexity refers to a fixed-window design as spatial size scales, not arbitrary global attention or every end-to-end operation. Benchmarks are source-specific.

Contributors

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction