From 0878dedb198d2ecde6ac8e0d7a8c6fa1766cba4f Mon Sep 17 00:00:00 2001 From: Ankle Date: Mon, 21 Sep 2026 07:22:28 +0000 Subject: [PATCH] readme --- README.md | 30 +++++++++++++++--------------- 1 file changed, 15 insertions(+), 15 deletions(-) diff --git a/README.md b/README.md index 23e494d..2143045 100644 --- a/README.md +++ b/README.md @@ -1,24 +1,24 @@ 这是张伟的本科毕业设计。基于yolov5s6模型，对其output进行了改进，增强yolov5s6的分类能力，使得这个模型能同时较好地完成“目标检测+分类”这两个任务。 -M0_yolov5_raw： -原生yolov5s6的output -![YOLOv5 Baseline](docs/images/1、YOLOv5-baseline.png) +M0_yolov5_raw：原生yolov5s6的output -G8_decoupled_head_light分支： -解耦output，将目标检测和分类任务的output拆开，便于后续单独增强分类的能力 -![YOLOv5s6 Decoupled Head](docs/images/2、YOLOv5s6-DecoupledHead.png) +![YOLOv5 Baseline](./docs/images/1、YOLOv5-baseline.png) -G11_cls_head_enhance_Lightweight分支: -在分类支添加两层卷积 -![YOLOv5s6 Head Variant](docs/images/3、YOLOv5s6-HeadVariant.png) +G8_decoupled_head_light分支：解耦output，将目标检测和分类任务的output拆开，便于后续单独增强分类的能力 -G13_decoupled_head_ECA分支： -在分类支添加ECA通道注意力机制 -![YOLOv5s6 Decoupled Head ECA](docs/images/4、YOLOv5s6-DecoupledHead-ECA.png) +![YOLOv5s6 Decoupled Head](./docs/images/2、YOLOv5s6-DecoupledHead.png) -G12_cls_head_enhance_and_ECA分支： -在分类支同时添加两层卷积和ECA -![YOLOv5s6 Head Variant ECA](docs/images/5、YOLOv5s6-HeadVariant-ECA.png) +G11_cls_head_enhance_Lightweight分支：在分类支添加两层卷积 + +![YOLOv5s6 Head Variant](./docs/images/3、YOLOv5s6-HeadVariant.png) + +G13_decoupled_head_ECA分支：在分类支添加ECA通道注意力机制 + +![YOLOv5s6 Decoupled Head ECA](./docs/images/4、YOLOv5s6-DecoupledHead-ECA.png) + +G12_cls_head_enhance_and_ECA分支：在分类支同时添加两层卷积和ECA + +![YOLOv5s6 Head Variant ECA](./docs/images/5、YOLOv5s6-HeadVariant-ECA.png) 以下是yolov5原git仓的readme