当前位置:网站首页>Record: yolov5 model pruning lightweight
Record: yolov5 model pruning lightweight
2022-07-19 05:46:00 【Just do it! ට⋆*】
Always wanted to try yolov5 Model pruning , Achieve lightweight effect , So I did the following :
thank cv Share with you , In case of accidents, record it yourself , Reference address
This paper uses pedestrian data set crowdhuman, More than 100 of them were extracted as training set tests , According to the operation instructions , The final result is as follows
yolov5s After pruning :
yolov5s Before pruning :
yolov5n Model :
边栏推荐
- PyTorch学习笔记【5】:使用卷积进行泛化
- Pointnet++代码详解(一):farthest_point_sample函数
- SQL time comparison
- 模型时间复杂度和空间复杂度
- Use ide to make jar package
- USB转TTL CH340模块安装(WIN10)
- Pointnet++代码详解(七):PointNetSetAbstractionMsg层
- Pointnet++ code details (II): Square_ Distance function
- 汉诺塔问题-->递归实现
- Pointnet++ code explanation (III): query_ ball_ Point function
猜你喜欢
随机推荐
USB转TTL CH340模块安装(WIN10)
9. Dim layer construction of data warehouse construction
运行基于MindSpore的yolov5流程记录
Pointnet++代码详解(五):sample_and_group函数和samle_and_group_all函数
Hanoi Tower problem -- > recursive implementation
C语言——冒泡排序
Pointnet++ code explanation (IV): index_ Points function
List and map
Pointnet++ code details (II): Square_ Distance function
JNI practical notes
【语音识别入门】基础概念与框架
Unable to determine Electron version. Please specify an Electron version
PCM silent detection
Time difference calculation
微信小程序中的WXML模板语法
The crowdhuman data set is processed according to the generated train Txt separate data set
Simple application of COAP in andorid
10. DWD layer construction of data warehouse construction
汉诺塔问题-->递归实现
Use of shutter Intl









