yolov8针对类别不平衡

针对类别不平衡，yolov8使用了一种平衡技术。以前的方法在训练过程中只针对COCO的部分类别进行训练，导致这些类别在推理过程中有更高的置信水平。然而，yolov8的方法是zero-shot的，它能够在不同类别之间实现平衡的置信水平，并预测出更多没有在COCO中出现的类别。这意味着yolov8能够更好地处理类别不平衡的情况，提高目标检测和实例分割的性能。<span class="em">1</span><span class="em">2</span><span class="em">3</span>
#### 引用[.reference_title]
- *1* *2* [论文解读：FastSAM | Fast Segment Anything | 基于yolov8-seg实现 比SAM快50倍](https://blog.csdn.net/a486259/article/details/131453794)[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_2"}}] [.reference_item style="max-width: 50%"]
- *3* [yolov8训练与应用](https://download.csdn.net/download/qq_34252622/87721371)[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_2"}}] [.reference_item style="max-width: 50%"]
[ .reference_list ]