# Hierarchical feature maps

Swin Transformer v2 is a vision Transformer model that achieves efficient image classification and dense recognition tasks through hierarchical feature maps and local window self-attention mechanisms.