README.md 2.97 KB
Newer Older
Your Name's avatar
Your Name committed
1
2
3
4
5
6
7
8
# YoloV5

## 模型介绍

YoloV5是一种单阶段目标检测算法,该算法在YOLOV4的基础上添加了一些新的改进思路,使其速度与精度都得到了极大的性能提升。

## 模型结构

Your Name's avatar
Your Name committed
9
YoloV5模型的主要改进思路有以下几点:
Your Name's avatar
Your Name committed
10
11
12
13
14
15

- 输入端的Mosaic数据增强、自适应锚框计算、自适应图像缩放操作;
- 主干网络的Focus结构与CSP结构;
- Neck端的FPN+PAN结构;
- 输出端的损失函数GIOU_Loss以及预测框筛选的DIOU_nms。

liucong's avatar
liucong committed
16
17
18
19
## python版本推理

下面介绍如何运行python代码示例,具体推理代码解析,在Doc/Tutorial_Python.md中有详细说明。

liucong's avatar
liucong committed
20
### 拉取镜像
Your Name's avatar
Your Name committed
21
22

在光源可拉取推理的docker镜像,YoloV5工程推荐的镜像如下:
Your Name's avatar
Your Name committed
23

Your Name's avatar
Your Name committed
24
25
26
```python
docker pull image.sourcefind.cn:5000/dcu/admin/base/custom:ort1.14.0_migraphx3.0.0-dtk22.10.1
```
Your Name's avatar
Your Name committed
27

liucong's avatar
liucong committed
28
29
30
31
32
33
### 推理示例

YoloV5模型的推理示例程序是YoloV5_infer_migraphx.py,使用如下命令运行该推理示例:

```
# 进入python示例目录
shizhm's avatar
shizhm committed
34
cd <path_to_yolov5_migraphx>/Python
liucong's avatar
liucong committed
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57

# 安装依赖
pip install -r requirements.txt

# 运行程序
python YoloV5_infer_migraphx.py \
	--imgpath 测试图像路径 \ 
	--modelpath onnx模型路径 \
	--objectThreshold 判断是否有物体阈值,默认0.5 \
	--confThreshold 置信度阈值,默认0.25 \
	--nmsThreshold nms阈值,默认0.5 \
```

程序运行结束会在当前目录生成YoloV5检测结果图像。

<img src="./Resource/Images/Result.jpg" alt="Result_2" style="zoom: 50%;" />

## C++版本推理

下面介绍如何运行C++代码示例,具体推理代码解析,在Doc/Tutorial_Cpp.md目录中有详细说明。

参考Python版本推理中的构建安装,在光源中拉取推理的docker镜像。

Your Name's avatar
Your Name committed
58
### 安装Opencv依赖
Your Name's avatar
Your Name committed
59

Your Name's avatar
Your Name committed
60
61
62
```python
cd <path_to_migraphx_samples>
sh ./3rdParty/InstallOpenCVDependences.sh
Your Name's avatar
Your Name committed
63
```
Your Name's avatar
Your Name committed
64
65
66
67
68
69
70
71
72
73
74

### 修改CMakeLists.txt

- 如果使用ubuntu系统,需要修改CMakeLists.txt中依赖库路径:
  将"${CMAKE_CURRENT_SOURCE_DIR}/depend/lib64/"修改为"${CMAKE_CURRENT_SOURCE_DIR}/depend/lib/"

- **MIGraphX2.3.0及以上版本需要c++17**


### 安装OpenCV并构建工程

Your Name's avatar
Your Name committed
75
```
Your Name's avatar
Your Name committed
76
77
78
79
rbuild build -d depend
```

### 设置环境变量
Your Name's avatar
Your Name committed
80

Your Name's avatar
Your Name committed
81
82
83
将依赖库依赖加入环境变量LD_LIBRARY_PATH,在~/.bashrc中添加如下语句:

**Centos**:
Your Name's avatar
Your Name committed
84
85

```
shizhm's avatar
shizhm committed
86
export LD_LIBRARY_PATH=<path_to_yolov5_migraphx>/depend/lib64/:$LD_LIBRARY_PATH
Your Name's avatar
Your Name committed
87
88
89
90
91
```

**Ubuntu**:

```
shizhm's avatar
shizhm committed
92
export LD_LIBRARY_PATH=<path_to_yolov5_migraphx>/depend/lib/:$LD_LIBRARY_PATH
Your Name's avatar
Your Name committed
93
94
```

Your Name's avatar
Your Name committed
95
96
97
98
99
100
然后执行:

```
source ~/.bashrc
```

liucong's avatar
liucong committed
101
### 推理示例
Your Name's avatar
Your Name committed
102

liucong's avatar
liucong committed
103
成功编译YoloV5工程后,执行如下命令运行该示例:
Your Name's avatar
Your Name committed
104
105

```
liucong's avatar
liucong committed
106
# 进入migraphx samples工程根目录
shizhm's avatar
shizhm committed
107
cd <path_to_yolov5_migraphx> 
Your Name's avatar
Your Name committed
108

liucong's avatar
liucong committed
109
110
# 进入build目录
cd ./build/
Your Name's avatar
Your Name committed
111

liucong's avatar
liucong committed
112
113
# 执行示例程序
./YOLOV5
Your Name's avatar
Your Name committed
114
```
Your Name's avatar
Your Name committed
115

liucong's avatar
liucong committed
116
程序运行结束会在build目录生成YoloV5检测结果图像。
Your Name's avatar
Your Name committed
117

liucong's avatar
liucong committed
118
<img src="./Resource/Images/Result.jpg" alt="Result" style="zoom:50%;" />
Your Name's avatar
Your Name committed
119
120
121
122
123
124
125
126

## 历史版本

​		https://developer.hpccube.com/codes/modelzoo/yolov5_migraphx

## 参考

​		https://github.com/ultralytics/yolov5