4. BatchNorm和NonLinearity 实现

爱自己没情敌丶

已于 2022-11-29 18:01:28 修改

阅读量682

点赞数

分类专栏： C++ 基于Eigen库实现CRN前向推理文章标签：算法深度学习

于 2022-11-29 17:38:37 首次发布

本文链接：https://blog.csdn.net/weixin_42216799/article/details/128102158

版权

本文档介绍了如何利用C++的Eigen库实现卷积神经网络（CNN）中的反卷积（TransposedConv2d）操作，包括带膨胀率的实现，并详细阐述了BatchNorm2d和Nonlinearity（如Sigmoid、Tanh、ReLU等）的实现过程，提供了相应的头文件和源代码示例。

摘要由CSDN通过智能技术生成

[C++ 基于Eigen库实现CRN前向推理]

第三部分：TransposedConv2d实现（含dilation）

1. 基于Eigen实现BatchNorm2d

1.1 Layer_BatchNorm2d.h

//
// Created by Koer on 2022/10/31.
//

#ifndef CRN_LAYER_BATCHNORM2D_H
#define CRN_LAYER_BATCHNORM2D_H


#include "Eigen"
#include "mat.h"
#include "Eigen/CXX11/Tensor"

class Layer_BatchNorm2d {
   
public:
    Layer_BatchNorm2d();

    Layer_BatchNorm2d(int64_t bn_ch);

    void LoadState(MATFile *pmFile, const std::string &state_preffix);

    void LoadTestState();

    Eigen::Tensor<float_t, 4> forward(Eigen::Tensor<float_t, 4> &input);

private:
    int64_t channels;
    Eigen::Tensor<float_t, 2> weights;
    Eigen::Tensor<float_t, 2> bias;
    Eigen::Tensor<float_t, 2> running_mean;
    Eigen::Tensor<float_t, 2> running_var;
    int32_t num_batches_tracked;


};


#endif //CRN_LAYER_BATCHNORM2D_H

1.2 Layer_BatchNorm2d.cpp

//
// Created by Koer on 2022/10/31.
//
#include "iostream"
#include "../include/Layer_BatchNorm2d.h"

Layer_BatchNorm2d::Layer_BatchNorm2d() {
   
    this->channels = 1;
}

Layer_BatchNorm2d::Layer_BatchNorm2d(int64_t bn_ch) {
   
    this->channels = bn_ch;
}

void Layer_BatchNorm2d::LoadState(MATFile *pmFile, const std::string &state_preffix) {
   
    std::string weight_name = state_preffix + "_weight";
    std::string bias_name = state_preffix + "_bias";
    std::string rm_name = state_preffix + "_running_mean";
    std::string rv_name = state_preffix + "_running_var";
    std::string nbt_name = state_preffix + "_num_batches_tracked";

    mxArray *pa = matGetVariable(pmFile, weight_name.c_str());
    auto *values = (float_t *) mxGetData(pa);
    long long dim1 = mxGetM(pa);
    long long dim2 = mxGetN(pa);
    this->weights.resize(dim1, dim2);
    int idx = 0;
    for (int i = 0; i < dim2; i++) {
   
        for (int j = 0; j < dim1; j++) {
   
            this->weights(j, i) =<