detr论文解读【End-to-End Object Detection with Transformers】

下载

论文地址
源码地址

论文解读

仅仅的目标检测部分进行解读

1、Background

detr论文解读【End-to-End Object Detection with Transformers】_第1张图片

2、Goal/Objective

detr论文解读【End-to-End Object Detection with Transformers】_第2张图片

3、Strategy

detr论文解读【End-to-End Object Detection with Transformers】_第3张图片

4、整体网络架构

detr论文解读【End-to-End Object Detection with Transformers】_第4张图片
detr论文解读【End-to-End Object Detection with Transformers】_第5张图片

detr论文解读【End-to-End Object Detection with Transformers】_第6张图片

5、Encoder完成的任务

detr论文解读【End-to-End Object Detection with Transformers】_第7张图片
detr论文解读【End-to-End Object Detection with Transformers】_第8张图片

6、优缺点

detr论文解读【End-to-End Object Detection with Transformers】_第9张图片

代码复现

详见 https://blog.csdn.net/weixin_44635198/article/details/128430111

你可能感兴趣的:(深度学习,#,transform,Python,目标检测,深度学习,人工智能)