senet融入yolov5
时间: 2023-09-10 15:10:29 浏览: 137
在YOLOv5中,作者将SENet融入了网络结构。SENet是一种用于深度卷积神经网络的注意力机制。它通过对输入特征图的通道维度进行自适应的注意力加权,来提高网络的表达能力。在YOLOv5中,作者在每个卷积层后面加入了SE结构,以增强网络对不同特征的关注能力。这种融合的方式使得YOLOv5在目标检测任务上取得了较好的性能。<span class="em">1</span><span class="em">2</span><span class="em">3</span>
#### 引用[.reference_title]
- *1* *3* [论文阅读—嵌入 SENet 结构的改进 YOLOV3 目标识别算法](https://blog.csdn.net/weixin_54723793/article/details/117751650)[target="_blank" data-report-click={"spm":"1018.2226.3001.9630","extra":{"utm_source":"vip_chatgpt_common_search_pc_result","utm_medium":"distribute.pc_search_result.none-task-cask-2~all~insert_cask~default-1-null.142^v93^chatsearchT3_1"}}] [.reference_item style="max-width: 50%"]
- *2* [主流注意力机制介绍以及如何添加到YOLOV5](https://blog.csdn.net/qq_41627642/article/details/127381209)[target="_blank" data-report-click={"spm":"1018.2226.3001.9630","extra":{"utm_source":"vip_chatgpt_common_search_pc_result","utm_medium":"distribute.pc_search_result.none-task-cask-2~all~insert_cask~default-1-null.142^v93^chatsearchT3_1"}}] [.reference_item style="max-width: 50%"]
[ .reference_list ]
阅读全文