yolo ghost
时间: 2023-10-19 17:32:21 浏览: 86
YOLO Ghost指的是在YOLOv8模型中使用GhostNetV2作为主干网络的一种配置。GhostNetV2是一种轻量级的卷积神经网络模型,具有较低的参数量和计算复杂度。与YOLOv8原来的主干网络DarkNet53相比,GhostNetV2的结构有较大差异。在实现YOLO Ghost时,我们使用了GhostNetV2的论文作者提供的轻量级版本,并针对YOLOv8模型特点进行了一定的优化。这种配置可以在保持较高检测精度的同时,减少了模型的计算和存储资源消耗。 如果您对GhostNetV2的详细信息感兴趣,可以查阅该论文的地址:https://arxiv.org/pdf/1911.11907v1.pdf。<span class="em">1</span><span class="em">2</span><span class="em">3</span>
#### 引用[.reference_title]
- *1* [YOLOv8主干网络升级篇:使用GhostNetV2增强模型性能](https://blog.csdn.net/wellcoder/article/details/131014868)[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: 33.333333333333336%"]
- *2* [YOLOV5的多主干网络(backbone)实现.zip](https://download.csdn.net/download/cuihao1995/83477139)[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: 33.333333333333336%"]
- *3* [目标检测算法——YOLOv5/YOLOv7改进结合轻量型Ghost模块](https://blog.csdn.net/m0_53578855/article/details/124056682)[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: 33.333333333333336%"]
[ .reference_list ]
阅读全文