# Swin Tranformer ### 應用到影像上的兩大挑戰: - large variations in the scale of visual entities - high resolution of pixels in images compared to words in text --- ## 使用方法: - hierarchical - shifted windows ![](https://i.imgur.com/9k1xPJD.png) --- ![](https://i.imgur.com/Q7g7bFN.png) --- ![](https://i.imgur.com/4hxzS4M.png) --- ![](https://i.imgur.com/OJcLGmq.png) --- ![](https://i.imgur.com/qP41Dsg.png) --- ![](https://i.imgur.com/NxU1d0R.png) --- ## object detection ![](https://i.imgur.com/quXToyl.png) --- ## semantic segmentation ![](https://i.imgur.com/rpemdGI.png) ---
{"title":"Swin Transformer","description":"type: slide","contributors":"[{\"id\":\"4724c2c9-74e7-418c-87c3-79a619b97461\",\"add\":634,\"del\":13}]"}
    101 views