Swin Transformer: Hierarchical Vision Transformer using Shifted Windows

TLDR

Challenges in adapting Transformer from language to vision arise from differences between the two domains, such as large variations in the scale of visual entities and the high resolution of pixels in images compared to words in text. The paper introduces Swin Transformer, a general‑purpose vision backbone that addresses domain differences by proposing a hierarchical Transformer with shifted windows. Swin Transformer employs a hierarchical architecture with shifted windows, limiting self‑attention to local windows for efficiency while enabling cross‑window connections, and achieving linear computational complexity with respect to image size. Swin Transformer attains state‑of‑the‑art performance on ImageNet‑1K, COCO, and ADE20K, surpassing prior methods by +2.7 box AP, +2.6 mask AP, and +3.2 mIoU, and its hierarchical shifted‑window design also benefits all‑MLP models. The code and models are publicly available at https://github.com/microsoft/Swin-Transformer.

Abstract

This paper presents a new vision Transformer, called Swin Transformer, that capably serves as a general-purpose backbone for computer vision. Challenges in adapting Transformer from language to vision arise from differences between the two domains, such as large variations in the scale of visual entities and the high resolution of pixels in images compared to words in text. To address these differences, we propose a hierarchical Transformer whose representation is computed with Shifted windows. The shifted windowing scheme brings greater efficiency by limiting self-attention computation to non-overlapping local windows while also allowing for cross-window connection. This hierarchical architecture has the flexibility to model at various scales and has linear computational complexity with respect to image size. These qualities of Swin Transformer make it compatible with a broad range of vision tasks, including image classification (87.3 top-1 accuracy on ImageNet-1K) and dense prediction tasks such as object detection (58.7 box AP and 51.1 mask AP on COCO test-dev) and semantic segmentation (53.5 mIoU on ADE20K val). Its performance surpasses the previous state-of-the-art by a large margin of +2.7 box AP and +2.6 mask AP on COCO, and +3.2 mIoU on ADE20K, demonstrating the potential of Transformer-based models as vision backbones. The hierarchical design and the shifted window approach also prove beneficial for all-MLP architectures. The code and models are publicly available at https://github.com/microsoft/Swin-Transformer.