当前位置:网站首页>Grouping convolution and DW convolution, residuals and inverted residuals, bottleneck and linearbottleneck
Grouping convolution and DW convolution, residuals and inverted residuals, bottleneck and linearbottleneck
2022-07-06 06:26:00 【jq_ ninety-eight】
Grouping convolution (Group Convolution)
Group convolution in ResNext Used in the
First of all, it must be clear :
Conventional convolution (Convolution) The parameter quantity of is :
K*K*C_in*n
K It's the size of the convolution kernel ,C_in yes input Of channel Count ,n Is the number of convolution kernels (output Of channel Count )
The parameter quantity of block convolution is :
K*K*C_in*n*1/g
K It's the size of the convolution kernel ,C_in yes input Of channel Count ,n Is the number of convolution kernels (output Of channel Count ),g Is the number of groups
DW(Depthwise Separable Conv)+PW(Pointwise Conv) Convolution
DW Convolution is also called deep separable convolution ,DW+PW The combination of MobileNet Used in
DW The parameter quantity of convolution is :
K*K*C_in ( here C_in = n)
K It's the size of the convolution kernel ,C_in yes input Of channel Count ,DW The convolution , The number of convolution kernels and input Of channel The same number
PW The parameter quantity of convolution is :
1*1*C_in*n
PW The convolution kernel of convolution is 1*1 size ,C_in yes input Of channel Count ,n Is the number of convolution kernels (output Of channel Count )
summary
- The parameter quantity of block convolution is conventional convolution (Convolution) Parameter quantity 1/g, among g Is the number of groups
- DW The parameter quantity of convolution is conventional convolution (Convolution) Parameter quantity 1/n, among n Is the number of convolution kernels
- When in packet convolution g=C_in, n=C_in when ,DW== Grouping convolution
Residuals And Inverted Residuals
bottleneck And linearbottleneck
Bottleneck It refers to the bottleneck layer ,Bottleneck Structure is actually to reduce the number of parameters ,Bottleneck Three steps are first PW Dimensionality reduction of data , Then the convolution of conventional convolution kernel , Last PW Dimension upgrading of data ( Similar to the hourglass ).
The focus here is on health in the network ( l ) dimension -> Convolution -> l ( drop ) Dimensional structure , Rather than shortcut
Linear Bottlececk: in the light of MobileNet v2 Medium Inverted residual block The last of the structure 1*1 The convolution layer uses a linear activation function , instead of relu Activation function
边栏推荐
- 基於JEECG-BOOT的list頁面的地址欄參數傳遞
- 数据库隔离级别
- Cannot create PoolableConnectionFactory (Could not create connection to database server. 错误
- [mqtt from getting started to improving series | 01] quickly build an mqtt test environment from 0 to 1
- Delete the variables added to watch1 in keil MDK
- Still worrying about how to write web automation test cases? Senior test engineers teach you selenium test case writing hand in hand
- Simulation volume leetcode [general] 1447 Simplest fraction
- 黑猫带你学UFS协议第4篇:UFS协议栈详解
- An article was uncovered to test the truth of outsourcing companies
- LeetCode 1200. 最小绝对差
猜你喜欢
随机推荐
Redis 核心技术与实战之 基本架构:一个键值数据库包含什么?
基於JEECG-BOOT的list頁面的地址欄參數傳遞
[C language] string left rotation
F - True Liars (种类并查集+DP)
模拟卷Leetcode【普通】1218. 最长定差子序列
Testing of web interface elements
Full link voltage measurement: building three models
Cannot create PoolableConnectionFactory (Could not create connection to database server. 错误
MFC on the conversion and display of long string unsigned char and CString
Summary of anomaly detection methods
Isam2 operation process
Simulation volume leetcode [general] 1062 Longest repeating substring
E - 食物链
On weak network test of special test
(中)苹果有开源,但又怎样呢?
Remember the implementation of a relatively complex addition, deletion and modification function based on jeecg-boot
Luogu p2141 abacus mental arithmetic test
Technology sharing | common interface protocol analysis
LeetCode 731. My schedule II
国际经贸合同翻译 中译英怎样效果好