本文主要解析caffe源码文件/src/caffe/layers/Conv_layer.cpp和Base_conv_layer.cpp，这两个文件主要实现卷积层的定义。
卷积神经网络（CNN）最核心的一个算法就是卷积运算，caffe框架在Conv_layer.cpp中实现了具体的卷积运算。Conv_layer.cpp中的类conv_layer并非是最基础的卷积层，它是继承了base_conv_layer，这个才是最基础的卷积层，很多卷积函数实际上是在这里面实现的。base_conv_layer又继承了基类layer。基类源码解析可见：：：http://blog.csdn.net/lanxuecc/article/details/53023211。
先来看看BaseConvolutionLayer中类的实现。
Base_conv_layer.hpp：：：：：：：
#ifndef CAFFE_BASE_CONVOLUTION_LAYER_HPP_
#define CAFFE_BASE_CONVOLUTION_LAYER_HPP_
#include <vector>
#include "caffe/blob.hpp"
#include "caffe/layer.hpp"
#include "caffe/proto/caffe.pb.h"
#include "caffe/util/im2col.hpp"
namespace caffe {
/**
* @brief Abstract base class that factors out the BLAS code common to
* ConvolutionLayer and DeconvolutionLayer.
*/
/*类BaseConvolutionLayer继承于类Layer*/
template <typename Dtype>
class BaseConvolutionLayer : public Layer<Dtype> {
public:
/*构造函数，实现之前先实现Layer的构造函数*/
explicit BaseConvolutionLayer(const LayerParameter& param)
: Layer<Dtype>(param) {}
/*层设置函数*/
virtual void LayerSetUp(const vector<Blob<Dtype>*>& bottom,
const vector<Blob<Dtype>*>& top);
/*内存分配与数据形状reshape函数*/
virtual void Reshape(const vector<Blob<Dtype>*>& bottom,
const vector<Blob<Dtype>*>& top);
/*返回最少的输入blob个数*/
virtual inline int MinBottomBlobs() const { return 1; }
/*返回最多的输出blob个数*/
virtual inline int MinTopBlobs() const { return 1; }
/*判断输入与输出blob个数是否一致*/
virtual inline bool EqualNumBottomTopBlobs() const { return true; }
protected:
// Helper functions that abstract away the column buffer and gemm arguments.
// The last argument in forward_cpu_gemm is so that we can skip the im2col if
// we just called weight_cpu_gemm with the same input.
// cpu实现前向传播的卷积操作
void forward_cpu_gemm(const Dtype* input, const Dtype* weights,
Dtype* output, bool skip_im2col = false);
// cpu实现前向传播的卷积操作后加上bias
void forward_cpu_bias(Dtype* output, const Dtype* bias);
// 后向传播求数据导数
void backward_cpu_gemm(const Dtype* input, const Dtype* weights,
Dtype* output);
// 后向传播求权重导数
void weight_cpu_gemm(const Dtype* input, const Dtype* output, Dtype*
weights);
// 后传播求bias（偏置项）导数
void backward_cpu_bias(Dtype* bias, const Dtype* input);
//上述5个函数的gpu实现
#ifndef CPU_ONLY
void forward_gpu_gemm(const Dtype* col_input, const Dtype* weights,
Dtype* output, bool skip_im2col = false);
void forward_gpu_bias(Dtype* output, const Dtype* bias);
void backward_gpu_gemm(const Dtype* input, const Dtype* weights,
Dtype* col_output);
void weight_gpu_gemm(const Dtype* col_input, const Dtype* output, Dtype*
weights);
void backward_gpu_bias(Dtype* bias, const Dtype* input);
#endif
/// @brief The spatial dimensions of the input.
inline int input_shape(int i) {
return (*bottom_shape_)[channel_axis_ + i];
}
// reverse_dimensions should return true iff we are implementing deconv, so
// that conv helpers know which dimensions are which.
virtual bool reverse_dimensions() = 0;
// Compute height_out_ and width_out_ from other parameters.
virtual void compute_output_shape() = 0;
/// @brief The spatial dimensions of a filter kernel.
/// 卷积核的形状 = [kernel_h, kernel_w]
Blob<int> kernel_shape_;
/// @brief The spatial dimensions of the stride.
/// 步长形状 = [stride_h, stride_w]
Blob<int> stride_;
/// @brief The spatial dimensions of the padding.
/// pad的形状 = [pad_h, pad_w]
Blob<int> pad_;
/// @brief The spatial dimensions of the dilation.
///
Blob<int> dilation_;
/// @brief The spatial dimensions of the convolution input.
/// 卷积的输入形状 = [输入图像通道数, 输入图像h, 输入图像w]
Blob<int> conv_input_shape_;
/// @brief The spatial dimensions of the col_buffer.
/// col_buffer的形状 = [kernel_dim_, conv_out_spatial_dim_]
vector<int> col_buffer_shape_;
/// @brief The spatial dimensions of the output.
/// 输出的形状
vector<int> output_shape_;
/// 输入的形状
const vector<int>* bottom_shape_;
/// 空间轴个数
int num_spatial_axes_;
/// 输入的维度 = 输入图像通道数*输入图像的h*输入图像w
int bottom_dim_;
// 输出维度 = 输出通道数*输出h*输出w
int top_dim_;
// 输入图像的第几个轴是通道
int channel_axis_;
// batchsize (一次处理的图片个数)
int num_;
// 图像的通道数
int channels_;
// 卷积组的大小（有多少个卷积）
int group_;
// 输出空间维度 = 卷积之后的图像长*卷积之后图像的宽
int out_spatial_dim_;
// 使用卷积组时用到
int weight_offset_;
// 卷积后的图像的通道数
int num_output_;
// 是否启用偏置
bool bias_term_;
// 是不是1x1卷积
bool is_1x1_;
// 强制使用n维通用卷积
bool force_nd_im2col_;
private:
// wrap im2col/col2im so we don't have to remember the (long) argument lists
// cpu代码将卷积核在图像上的滑动转换为了矩阵
inline void conv_im2col_cpu(const Dtype* data, Dtype* col_buff) {
if (!force_nd_im2col_ && num_spatial_axes_ == 2) {
im2col_cpu(data, conv_in_channels_,
conv_input_shape_.cpu_data()[1], conv_input_shape_.cpu_data()[2],
kernel_shape_.cpu_data()[0], kernel_shape_.cpu_data()[1],
pad_.cpu_data()[0], pad_.cpu_data()[1],
stride_.cpu_data()[0], stride_.cpu_data()[1],
dilation_.cpu_data()[0], dilation_.cpu_data()[1], col_buff);
} else {
im2col_nd_cpu(data, num_spatial_axes_, conv_input_shape_.cpu_data(),
col_buffer_shape_.data(), kernel_shape_.cpu_data(),
pad_.cpu_data(), stride_.cpu_data(), dilation_.cpu_data(), col_buff);
}
}
// cpu代码将矩阵转换成为图像
inline void conv_col2im_cpu(const Dtype* col_buff, Dtype* data) {
if (!force_nd_im2col_ && num_spatial_axes_ == 2) {
col2im_cpu(col_buff, conv_in_channels_,
conv_input_shape_.cpu_data()[1], conv_input_shape_.cpu_data()[2],
kernel_shape_.cpu_data()[0], kernel_shape_.cpu_data()[1],
pad_.cpu_data()[0], pad_.cpu_data()[1],
stride_.cpu_data()[0], stride_.cpu_data()[1],
dilation_.cpu_data()[0], dilation_.cpu_data()[1], data);
} else {
col2im_nd_cpu(col_buff, num_spatial_axes_, conv_input_shape_.cpu_data(),
col_buffer_shape_.data(), kernel_shape_.cpu_data(),
pad_.cpu_data(), stride_.cpu_data(), dilation_.cpu_data(), data);
}
}
/*gpu代码实现上述两个函数*/
#ifndef CPU_ONLY
inline void conv_im2col_gpu(const Dtype* data, Dtype* col_buff) {
if (!force_nd_im2col_ && num_spatial_axes_ == 2) {
im2col_gpu(data, conv_in_channels_,
conv_input_shape_.cpu_data()[1], conv_input_shape_.cpu_data()[2],
kernel_shape_.cpu_data()[0], kernel_shape_.cpu_data()[1],
pad_.cpu_data()[0], pad_.cpu_data()[1],
stride_.cpu_data()[0], stride_.cpu_data()[1],
dilation_.cpu_data()[0], dilation_.cpu_data()[1], col_buff);
} else {
im2col_nd_gpu(data, num_spatial_axes_, num_kernels_im2col_,
conv_input_shape_.gpu_data(), col_buffer_.gpu_shape(),
kernel_shape_.gpu_data(), pad_.gpu_data(),
stride_.gpu_data(), dilation_.gpu_data(), col_buff);
}
}
inline void conv_col2im_gpu(const Dtype* col_buff, Dtype* data) {
if (!force_nd_im2col_ && num_spatial_axes_ == 2) {
col2im_gpu(col_buff, conv_in_channels_,
conv_input_shape_.cpu_data()[1], conv_input_shape_.cpu_data()[2],
kernel_shape_.cpu_data()[0], kernel_shape_.cpu_data()[1],
pad_.cpu_data()[0], pad_.cpu_data()[1],
stride_.cpu_data()[0], stride_.cpu_data()[1],
dilation_.cpu_data()[0], dilation_.cpu_data()[1], data);
} else {
col2im_nd_gpu(col_buff, num_spatial_axes_, num_kernels_col2im_,
conv_input_shape_.gpu_data(), col_buffer_.gpu_shape(),
kernel_shape_.gpu_data(), pad_.gpu_data(), stride_.gpu_data(),
dilation_.gpu_data(), data);
}
}
#endif
// conv_in_channels_ * conv_out_spatial_dim_
int num_kernels_im2col_;
// num_kernels_col2im_ = reverse_dimensions() ? top_dim_ : bottom_dim_
int num_kernels_col2im_;
// 卷积的输出通道数 ,在参数配置文件中设置
int conv_out_channels_;
// 卷积的输入通道数 （即输入图像的通道数）
int conv_in_channels_;
// 卷积的输出的空间维度 = 卷积后图像h*卷积后图像w
int conv_out_spatial_dim_;
// 卷积核的维度 = 输入图像的维度*卷积核的h*卷积核的w
int kernel_dim_;
// 在使用group参数的时候使用的offset
int col_offset_;
int output_offset_;
// im2col的时候使用的存储空间
Blob<Dtype> col_buffer_;
// 将偏置扩展成矩阵
Blob<Dtype> bias_multiplier_;
};
} // namespace caffe
#endif // CAFFE_BASE_CONVOLUTION_LAYER_HPP_Base_conv_layer.cpp：：：：：：：
#include <algorithm>
#include <vector>
#include "caffe/filler.hpp"
#include "caffe/layers/base_conv_layer.hpp"
#include "caffe/util/im2col.hpp"
#include "caffe/util/math_functions.hpp"
namespace caffe {
/*BaseConvolutionLayer类的LayerSetUp函数的实现，参数当前层的输入与输出的blob数据的向量*/
template <typename Dtype>
void BaseConvolutionLayer<Dtype>::LayerSetUp(const vector<Blob<Dtype>*>& bottom,
const vector<Blob<Dtype>*>& top) {
// Configure the kernel size, padding, stride, and inputs.
ConvolutionParameter conv_param = this->layer_param_.convolution_param(); //将传入的参数赋值进来
//im2col,一般情况下num_spatial_axes_==2,即将2维图像拉成向量，但force_nd_im2col_针对的是更general的情况n-d“图像”
force_nd_im2col_ = conv_param.force_nd_im2col();
channel_axis_ = bottom[0]->CanonicalAxisIndex(conv_param.axis());
const int first_spatial_axis = channel_axis_ + 1;
const int num_axes = bottom[0]->num_axes();
num_spatial_axes_ = num_axes - first_spatial_axis;
CHECK_GE(num_spatial_axes_, 0);
vector<int> bottom_dim_blob_shape(1, num_spatial_axes_ + 1);
//当num_spatial_axes_==2时，spatial_dim_blob_shape这个vector只包含一个元素且值为2
vector<int> spatial_dim_blob_shape(1, std::max(num_spatial_axes_, 1));
// Setup filter kernel dimensions (kernel_shape_).
// 以spatial_dim_blob_shape为参数来构造一个Blob，即kernel_shape_，
// 则这个Blob的维度信息只包含一个维度，值为2,也就是说这个Blob的count_==2。
// 尽管这个Blob的维度信息只包含一个维度，因为在后续的计算（Im2col）中，
// 我只关心这个Blob中的数据的值，而不关心这个Blob的shape信息，
// 例如在Im2col()中，只要取出相应数值即可kernel_shape_.cpu_data()[0], kernel_shape_.cpu_data()[1],pad_.cpu_data()[0], pad_.cpu_data()[1]。
kernel_shape_.Reshape(spatial_dim_blob_shape);
int* kernel_shape_data = kernel_shape_.mutable_cpu_data();
if (conv_param.has_kernel_h() || conv_param.has_kernel_w()) {
CHECK_EQ(num_spatial_axes_, 2)
<< "kernel_h & kernel_w can only be used for 2D convolution.";
CHECK_EQ(0, conv_param.kernel_size_size())
<< "Either kernel_size or kernel_h/w should be specified; not both.";
kernel_shape_data[0] = conv_param.kernel_h();//kernel_shape_data是一个二维数组
kernel_shape_data[1] = conv_param.kernel_w();
} else {
const int num_kernel_dims = conv_param.kernel_size_size();
CHECK(num_kernel_dims == 1 || num_kernel_dims == num_spatial_axes_)
<< "kernel_size must be specified once, or once per spatial dimension "
<< "(kernel_size specified " << num_kernel_dims << " times; "
<< num_spatial_axes_ << " spatial dims).";
for (int i = 0; i < num_spatial_axes_; ++i) {
kernel_shape_data[i] =
conv_param.kernel_size((num_kernel_dims == 1) ? 0 : i);
}
}
for (int i = 0; i < num_spatial_axes_; ++i) {
CHECK_GT(kernel_shape_data[i], 0) << "Filter dimensions must be nonzero.";
}
// Setup stride dimensions (stride_).
stride_.Reshape(spatial_dim_blob_shape);
int* stride_data = stride_.mutable_cpu_data();
if (conv_param.has_stride_h() || conv_param.has_stride_w()) {
CHECK_EQ(num_spatial_axes_, 2)
<< "stride_h & stride_w can only be used for 2D convolution.";
CHECK_EQ(0, conv_param.stride_size())
<< "Either stride or stride_h/w should be specified; not both.";
stride_data[0] = conv_param.stride_h();
stride_data[1] = conv_param.stride_w();
} else {
const int num_stride_dims = conv_param.stride_size();
CHECK(num_stride_dims == 0 || num_stride_dims == 1 ||
num_stride_dims == num_spatial_axes_)
<< "stride must be specified once, or once per spatial dimension "
<< "(stride specified " << num_stride_dims << " times; "
<< num_spatial_axes_ << " spatial dims).";
const int kDefaultStride = 1;
for (int i = 0; i < num_spatial_axes_; ++i) {
stride_data[i] = (num_stride_dims == 0) ? kDefaultStride :
conv_param.stride((num_stride_dims == 1) ? 0 : i);
CHECK_GT(stride_data[i], 0) << "Stride dimensions must be nonzero.";
}
}
// Setup pad dimensions (pad_).
pad_.Reshape(spatial_dim_blob_shape);
int* pad_data = pad_.mutable_cpu_data();
if (conv_param.has_pad_h() || conv_param.has_pad_w()) {
CHECK_EQ(num_spatial_axes_, 2)
<< "pad_h & pad_w can only be used for 2D convolution.";
CHECK_EQ(0, conv_param.pad_size())
<< "Either pad or pad_h/w should be specified; not both.";
pad_data[0] = conv_param.pad_h();
pad_data[1] = conv_param.pad_w();
} else {
const int num_pad_dims = conv_param.pad_size();
CHECK(num_pad_dims == 0 || num_pad_dims == 1 ||
num_pad_dims == num_spatial_axes_)
<< "pad must be specified once, or once per spatial dimension "
<< "(pad specified " << num_pad_dims << " times; "
<< num_spatial_axes_ << " spatial dims).";
const int kDefaultPad = 0;
for (int i = 0; i < num_spatial_axes_; ++i) {
pad_data[i] = (num_pad_dims == 0) ? kDefaultPad :
conv_param.pad((num_pad_dims == 1) ? 0 : i);
}
}
// Setup dilation dimensions (dilation_).
dilation_.Reshape(spatial_dim_blob_shape);
int* dilation_data = dilation_.mutable_cpu_data();
const int num_dilation_dims = conv_param.dilation_size();
CHECK(num_dilation_dims == 0 || num_dilation_dims == 1 ||
num_dilation_dims == num_spatial_axes_)
<< "dilation must be specified once, or once per spatial dimension "
<< "(dilation specified " << num_dilation_dims << " times; "
<< num_spatial_axes_ << " spatial dims).";
const int kDefaultDilation = 1;
for (int i = 0; i < num_spatial_axes_; ++i) {
dilation_data[i] = (num_dilation_dims == 0) ? kDefaultDilation :
conv_param.dilation((num_dilation_dims == 1) ? 0 : i);
}
// Special case: im2col is the identity for 1x1 convolution with stride 1
// and no padding, so flag for skipping the buffer and transformation.
is_1x1_ = true;
for (int i = 0; i < num_spatial_axes_; ++i) {
is_1x1_ &=
kernel_shape_data[i] == 1 && stride_data[i] == 1 && pad_data[i] == 0;
if (!is_1x1_) { break; }
}
// Configure output channels and groups.
channels_ = bottom[0]->shape(channel_axis_);
num_output_ = this->layer_param_.convolution_param().num_output();
CHECK_GT(num_output_, 0);
group_ = this->layer_param_.convolution_param().group();
CHECK_EQ(channels_ % group_, 0);
CHECK_EQ(num_output_ % group_, 0)
<< "Number of output should be multiples of group.";
if (reverse_dimensions()) {
conv_out_channels_ = channels_;
conv_in_channels_ = num_output_;
} else {
conv_out_channels_ = num_output_;
conv_in_channels_ = channels_;
}
// Handle the parameters: weights and biases.
// - blobs_[0] holds the filter weights
// - blobs_[1] holds the biases (optional)
vector<int> weight_shape(2);
weight_shape[0] = conv_out_channels_;
weight_shape[1] = conv_in_channels_ / group_;
for (int i = 0; i < num_spatial_axes_; ++i) {
weight_shape.push_back(kernel_shape_data[i]);
}
bias_term_ = this->layer_param_.convolution_param().bias_term();
vector<int> bias_shape(bias_term_, num_output_);
if (this->blobs_.size() > 0) {
CHECK_EQ(1 + bias_term_, this->blobs_.size())
<< "Incorrect number of weight blobs.";
if (weight_shape != this->blobs_[0]->shape()) {
Blob<Dtype> weight_shaped_blob(weight_shape);
LOG(FATAL) << "Incorrect weight shape: expected shape "
<< weight_shaped_blob.shape_string() << "; instead, shape was "
<< this->blobs_[0]->shape_string();
}
if (bias_term_ && bias_shape != this->blobs_[1]->shape()) {
Blob<Dtype> bias_shaped_blob(bias_shape);
LOG(FATAL) << "Incorrect bias shape: expected shape "
<< bias_shaped_blob.shape_string() << "; instead, shape was "
<< this->blobs_[1]->shape_string();
}
LOG(INFO) << "Skipping parameter initialization";
} else {
if (bias_term_) {
this->blobs_.resize(2);
} else {
this->blobs_.resize(1);
}
// Initialize and fill the weights:
// output channels x input channels per-group x kernel height x kernel width
this->blobs_[0].reset(new Blob<Dtype>(weight_shape));//blobs_[0]的维度信息是四个维度，count_为四个维度的值相乘
shared_ptr<Filler<Dtype> > weight_filler(GetFiller<Dtype>(
this->layer_param_.convolution_param().weight_filler()));
weight_filler->Fill(this->blobs_[0].get());
// If necessary, initialize and fill the biases.
if (bias_term_) {
this->blobs_[1].reset(new Blob<Dtype>(bias_shape));//blobs_[1]的维度信息是1个维度，count_为num_output_
shared_ptr<Filler<Dtype> > bias_filler(GetFiller<Dtype>(
this->layer_param_.convolution_param().bias_filler()));
bias_filler->Fill(this->blobs_[1].get());
}
}
kernel_dim_ = this->blobs_[0]->count(1);
//写成(conv_out_channels_ / group_) * kernel_dim_更直观。这个offset是相对group分组来讲的。
weight_offset_ = conv_out_channels_ * kernel_dim_ / group_;
// Propagate gradients to the parameters (as directed by backward pass).
this->param_propagate_down_.resize(this->blobs_.size(), true);
}
template <typename Dtype>
void BaseConvolutionLayer<Dtype>::Reshape(const vector<Blob<Dtype>*>& bottom,
const vector<Blob<Dtype>*>& top) {
const int first_spatial_axis = channel_axis_ + 1;
CHECK_EQ(bottom[0]->num_axes(), first_spatial_axis + num_spatial_axes_)
<< "bottom num_axes may not change.";
num_ = bottom[0]->count(0, channel_axis_);
CHECK_EQ(bottom[0]->shape(channel_axis_), channels_)
<< "Input size incompatible with convolution kernel.";
// TODO: generalize to handle inputs of different shapes.
for (int bottom_id = 1; bottom_id < bottom.size(); ++bottom_id) {
CHECK(bottom[0]->shape() == bottom[bottom_id]->shape())
<< "All inputs must have the same shape.";
}
// Shape the tops.
// 卷积层应该默认有多少个bottom 就有多少个top输出
bottom_shape_ = &bottom[0]->shape();
compute_output_shape();
vector<int> top_shape(bottom[0]->shape().begin(),
bottom[0]->shape().begin() + channel_axis_);
top_shape.push_back(num_output_);
for (int i = 0; i < num_spatial_axes_; ++i) {
top_shape.push_back(output_shape_[i]);
}
for (int top_id = 0; top_id < top.size(); ++top_id) {
top[top_id]->Reshape(top_shape);
}
if (reverse_dimensions()) {
conv_out_spatial_dim_ = bottom[0]->count(first_spatial_axis);
} else {
conv_out_spatial_dim_ = top[0]->count(first_spatial_axis);
}
//group分，对conv_in_channels分组; 卷积窗口在输入“图像”上按步长滑动，
//（可以想象）形成了多个子图;然后将所有子图拉成一列，列的长度就是col_offset_。
col_offset_ = kernel_dim_ * conv_out_spatial_dim_;
//卷积层的输出特征图也要分组，当然group_默认为1。
//写成(conv_out_channels_ / group_) * conv_out_spatial_dim_更直观
output_offset_ = conv_out_channels_ * conv_out_spatial_dim_ / group_;
// Setup input dimensions (conv_input_shape_).
vector<int> bottom_dim_blob_shape(1, num_spatial_axes_ + 1);
conv_input_shape_.Reshape(bottom_dim_blob_shape);
int* conv_input_shape_data = conv_input_shape_.mutable_cpu_data();
for (int i = 0; i < num_spatial_axes_ + 1; ++i) {
if (reverse_dimensions()) {
conv_input_shape_data[i] = top[0]->shape(channel_axis_ + i);
} else {
conv_input_shape_data[i] = bottom[0]->shape(channel_axis_ + i);
}
}
// The im2col result buffer will only hold one image at a time to avoid
// overly large memory usage. In the special case of 1x1 convolution
// it goes lazily unused to save memory.
col_buffer_shape_.clear();//col_buffer_shape_是一个vector
col_buffer_shape_.push_back(kernel_dim_ * group_);//所有conv_in_channels_个输入的channels都包含其中。
for (int i = 0; i < num_spatial_axes_; ++i) {
if (reverse_dimensions()) {
col_buffer_shape_.push_back(input_shape(i + 1));
} else {
col_buffer_shape_.push_back(output_shape_[i]);
}
}
//一般情况下，col_buffer_的维度信息为三个维度。
// col_buffer_shape_的存储的元素为：kernel_dim_ * group_，
// 输出特征图的H， 输出特征图的W。
//可以认为col_buffer_内所存储的数据的维度为：(kernel_dim_ * group_) × H × W，且与kernel_dim_ x conv_out_spatial_dim_有密切关系.
col_buffer_.Reshape(col_buffer_shape_);
bottom_dim_ = bottom[0]->count(channel_axis_);
top_dim_ = top[0]->count(channel_axis_);
num_kernels_im2col_ = conv_in_channels_ * conv_out_spatial_dim_;
num_kernels_col2im_ = reverse_dimensions() ? top_dim_ : bottom_dim_;
// Set up the all ones "bias multiplier" for adding biases by BLAS
out_spatial_dim_ = top[0]->count(first_spatial_axis);
if (bias_term_) {
vector<int> bias_multiplier_shape(1, out_spatial_dim_);
bias_multiplier_.Reshape(bias_multiplier_shape);//bias_multiplier_这个Blob的count_为out_spatial_dim_，是输出特征图的H×W
caffe_set(bias_multiplier_.count(), Dtype(1),
bias_multiplier_.mutable_cpu_data());
}
}
/*
只是对一张图像进行前向传播！
与全连接层类比，conv_out_channels_ / group_相当与全连接层的输出神经元个数;
conv_out_spatial_dim_相当于全连接层中的样本个数;
kernel_dim_相当与全连接层中每个样本特征向量的维数。
*/
//实现前向传播卷积操作
template <typename Dtype>
void BaseConvolutionLayer<Dtype>::forward_cpu_gemm(const Dtype* input,
const Dtype* weights, Dtype* output, bool skip_im2col) {
const Dtype* col_buff = input;
if (!is_1x1_) {
if (!skip_im2col) {
// 如果没有1x1卷积，也没有skip_im2col
// 则使用conv_im2col_cpu对使用卷积核滑动过程中的每一个kernel大小的图像块
// 变成一个列向量，形成一个height=kernel_dim_的
// width = 卷积后图像heght*卷积后图像width
conv_im2col_cpu(input, col_buffer_.mutable_cpu_data());
}
col_buff = col_buffer_.cpu_data();
}
// 使用caffe的cpu_gemm来进行计算
// 假设输入是20个feature map，输出是10个feature map，group_=2
// 那么他就会把这个训练网络分解成两个10->5的网络，由于两个网络结构是
// 一模一样的，那么就可以利用多个GPU完成训练加快训练速度
for (int g = 0; g < group_; ++g) {
caffe_cpu_gemm<Dtype>(CblasNoTrans, CblasNoTrans, conv_out_channels_ /
group_, conv_out_spatial_dim_, kernel_dim_,
(Dtype)1., weights + weight_offset_ * g, col_buff + col_offset_ * g,
(Dtype)0., output + output_offset_ * g);
//weights <--- blobs_[0]->cpu_data()。类比全连接层，
//weights为权重，col_buff相当与数据，矩阵相乘weights×col_buff.
//其中，weights的维度为(conv_out_channels_ /group_) x kernel_dim_，
//col_buff的维度为kernel_dim_ x conv_out_spatial_dim_，
//output的维度为(conv_out_channels_ /group_) x conv_out_spatial_dim_.
}
}
//前身传播卷积后加bias
template <typename Dtype>
void BaseConvolutionLayer<Dtype>::forward_cpu_bias(Dtype* output,
const Dtype* bias) {
// output = bias * bias_multiplier_
// num_output 与 conv_out_channel是一样的
// num_output_ X out_spatial_dim_ = num_output_ X 1 1 X out_spatial_dim_
caffe_cpu_gemm<Dtype>(CblasNoTrans, CblasNoTrans, num_output_,
out_spatial_dim_, 1, (Dtype)1., bias, bias_multiplier_.cpu_data(),
(Dtype)1., output);
}
//反向传播，计算关于bottom data的导数以便传给下一层
template <typename Dtype>
void BaseConvolutionLayer<Dtype>::backward_cpu_gemm(const Dtype* output,
const Dtype* weights, Dtype* input) {
Dtype* col_buff = col_buffer_.mutable_cpu_data();
if (is_1x1_) {
col_buff = input;
}
for (int g = 0; g < group_; ++g) {
//kernel_dim_=c_in*kernel_h*kernel_w
//conv_out_spatial_dim_=H_out*W_out
//conv_out_channels_=c_out
//weights是c_out*kernel_dim_
//output是c_out*conv_out_spatial_dim_
//col_buff是输入feature map的转换矩阵，kernel_dim_*conv_out_spatial_dim_
caffe_cpu_gemm<Dtype>(CblasTrans, CblasNoTrans, kernel_dim_,
conv_out_spatial_dim_, conv_out_channels_ / group_,
(Dtype)1., weights + weight_offset_ * g, output + output_offset_ * g,
(Dtype)0., col_buff + col_offset_ * g);
}
if (!is_1x1_) {
//转换为输入feature map
conv_col2im_cpu(col_buff, input);
}
}
//反向传播计算关于权重的导数用于更新权重
template <typename Dtype>
void BaseConvolutionLayer<Dtype>::weight_cpu_gemm(const Dtype* input,
const Dtype* output, Dtype* weights) {
const Dtype* col_buff = input;
if (!is_1x1_) {
conv_im2col_cpu(input, col_buffer_.mutable_cpu_data());
col_buff = col_buffer_.cpu_data();
}
for (int g = 0; g < group_; ++g) {
caffe_cpu_gemm<Dtype>(CblasNoTrans, CblasTrans, conv_out_channels_ / group_,
kernel_dim_, conv_out_spatial_dim_,
(Dtype)1., output + output_offset_ * g, col_buff + col_offset_ * g,
(Dtype)1., weights + weight_offset_ * g);
}
}
//反向传播计算关于bias的导数
template <typename Dtype>
void BaseConvolutionLayer<Dtype>::backward_cpu_bias(Dtype* bias,
const Dtype* input) {
caffe_cpu_gemv<Dtype>(CblasNoTrans, num_output_, out_spatial_dim_, 1.,
input, bias_multiplier_.cpu_data(), 1., bias);
}
/*用到gpu时的相应的处理方式*/
#ifndef CPU_ONLY
/*gpu前向传播卷积计算函数*/
template <typename Dtype>
void BaseConvolutionLayer<Dtype>::forward_gpu_gemm(const Dtype* input,
const Dtype* weights, Dtype* output, bool skip_im2col) {
const Dtype* col_buff = input;
if (!is_1x1_) {
if (!skip_im2col) {
conv_im2col_gpu(input, col_buffer_.mutable_gpu_data());
}
col_buff = col_buffer_.gpu_data();
}
for (int g = 0; g < group_; ++g) {
caffe_gpu_gemm<Dtype>(CblasNoTrans, CblasNoTrans, conv_out_channels_ /
group_, conv_out_spatial_dim_, kernel_dim_,
(Dtype)1., weights + weight_offset_ * g, col_buff + col_offset_ * g,
(Dtype)0., output + output_offset_ * g);
}
}
/*gpu前向传播卷积后加bias*/
template <typename Dtype>
void BaseConvolutionLayer<Dtype>::forward_gpu_bias(Dtype* output,
const Dtype* bias) {
caffe_gpu_gemm<Dtype>(CblasNoTrans, CblasNoTrans, num_output_,
out_spatial_dim_, 1, (Dtype)1., bias, bias_multiplier_.gpu_data(),
(Dtype)1., output);
}
/*gpu后向计算数据的导数*/
template <typename Dtype>
void BaseConvolutionLayer<Dtype>::backward_gpu_gemm(const Dtype* output,
const Dtype* weights, Dtype* input) {
Dtype* col_buff = col_buffer_.mutable_gpu_data();
if (is_1x1_) {
col_buff = input;
}
for (int g = 0; g < group_; ++g) {
caffe_gpu_gemm<Dtype>(CblasTrans, CblasNoTrans, kernel_dim_,
conv_out_spatial_dim_, conv_out_channels_ / group_,
(Dtype)1., weights + weight_offset_ * g, output + output_offset_ * g,
(Dtype)0., col_buff + col_offset_ * g);
}
if (!is_1x1_) {
conv_col2im_gpu(col_buff, input);
}
}
/*gpu后向传播计算权重导数以更新*/
template <typename Dtype>
void BaseConvolutionLayer<Dtype>::weight_gpu_gemm(const Dtype* input,
const Dtype* output, Dtype* weights) {
const Dtype* col_buff = input;
if (!is_1x1_) {
conv_im2col_gpu(input, col_buffer_.mutable_gpu_data());
col_buff = col_buffer_.gpu_data();
}
for (int g = 0; g < group_; ++g) {
caffe_gpu_gemm<Dtype>(CblasNoTrans, CblasTrans, conv_out_channels_ / group_,
kernel_dim_, conv_out_spatial_dim_,
(Dtype)1., output + output_offset_ * g, col_buff + col_offset_ * g,
(Dtype)1., weights + weight_offset_ * g);
}
}
/*gpu后向传播以计算bias导数*/
template <typename Dtype>
void BaseConvolutionLayer<Dtype>::backward_gpu_bias(Dtype* bias,
const Dtype* input) {
caffe_gpu_gemv<Dtype>(CblasNoTrans, num_output_, out_spatial_dim_, 1.,
input, bias_multiplier_.gpu_data(), 1., bias);
}
#endif // !CPU_ONLY
INSTANTIATE_CLASS(BaseConvolutionLayer);
} // namespace caffe
再来看看Conv_layer类的具体实现
Conv_layer.hpp：：：
#ifndef CAFFE_CONV_LAYER_HPP_
#define CAFFE_CONV_LAYER_HPP_
#include <vector>
#include "caffe/blob.hpp"
#include "caffe/layer.hpp"
#include "caffe/proto/caffe.pb.h"
#include "caffe/layers/base_conv_layer.hpp"
namespace caffe {
/**
* @brief Convolves the input image with a bank of learned filters,
* and (optionally) adds biases.
*
* Caffe convolves by reduction to matrix multiplication. This achieves
* high-throughput and generality of input and filter dimensions but comes at
* the cost of memory for matrices. This makes use of efficiency in BLAS.
*
* The input is "im2col" transformed to a channel K' x H x W data matrix
* for multiplication with the N x K' x H x W filter matrix to yield a
* N' x H x W output matrix that is then "col2im" restored. K' is the
* input channel * kernel height * kernel width dimension of the unrolled
* inputs so that the im2col matrix has a column for each input region to
* be filtered. col2im restores the output spatial structure by rolling up
* the output channel N' columns of the output matrix.
*/
template <typename Dtype> /*卷积层layer*/
class ConvolutionLayer : public BaseConvolutionLayer<Dtype> {
public:
/**
* @param param provides ConvolutionParameter convolution_param,
* with ConvolutionLayer options:
* - num_output. The number of filters.
* - kernel_size / kernel_h / kernel_w. The filter dimensions, given by
* kernel_size for square filters or kernel_h and kernel_w for rectangular
* filters.
* - stride / stride_h / stride_w (\b optional, default 1). The filter
* stride, given by stride_size for equal dimensions or stride_h and stride_w
* for different strides. By default the convolution is dense with stride 1.
* - pad / pad_h / pad_w (\b optional, default 0). The zero-padding for
* convolution, given by pad for equal dimensions or pad_h and pad_w for
* different padding. Input padding is computed implicitly instead of
* actually padding.
* - dilation (\b optional, default 1). The filter
* dilation, given by dilation_size for equal dimensions for different
* dilation. By default the convolution has dilation 1.
* - group (\b optional, default 1). The number of filter groups. Group
* convolution is a method for reducing parameterization by selectively
* connecting input and output channels. The input and output channel dimensions must be divisible
* by the number of groups. For group @f$ \geq 1 @f$, the
* convolutional filters' input and output channels are separated s.t. each
* group takes 1 / group of the input channels and makes 1 / group of the
* output channels. Concretely 4 input channels, 8 output channels, and
* 2 groups separate input channels 1-2 and output channels 1-4 into the
* first group and input channels 3-4 and output channels 5-8 into the second
* group.
* - bias_term (\b optional, default true). Whether to have a bias.
* - engine: convolution has CAFFE (matrix multiplication) and CUDNN (library
* kernels + stream parallelism) engines.
*/
/*构造函数*/
explicit ConvolutionLayer(const LayerParameter& param)
: BaseConvolutionLayer<Dtype>(param) {}
/*返回当前层的类型*/
virtual inline const char* type() const { return "Convolution"; }
protected:
/*cpu中执行的前向传播函数*/
virtual void Forward_cpu(const vector<Blob<Dtype>*>& bottom,
const vector<Blob<Dtype>*>& top);
/*gpu中执行的前向传播函数*/
virtual void Forward_gpu(const vector<Blob<Dtype>*>& bottom,
const vector<Blob<Dtype>*>& top);
/*cpu中执行的返向传播函数*/
virtual void Backward_cpu(const vector<Blob<Dtype>*>& top,
const vector<bool>& propagate_down, const vector<Blob<Dtype>*>& bottom);
/*cpu中执行的返向传播函数*/
virtual void Backward_gpu(const vector<Blob<Dtype>*>& top,
const vector<bool>& propagate_down, const vector<Blob<Dtype>*>& bottom);
virtual inline bool reverse_dimensions() { return false; }
/*计算输出数据的blob形状*/
virtual void compute_output_shape();
};
} // namespace caffe
#endif // CAFFE_CONV_LAYER_HPP_
Conv_layer.cpp：：：
#include <vector>
#include "caffe/layers/conv_layer.hpp"
namespace caffe {
template <typename Dtype>
void ConvolutionLayer<Dtype>::compute_output_shape() {
const int* kernel_shape_data = this->kernel_shape_.cpu_data();
const int* stride_data = this->stride_.cpu_data();
const int* pad_data = this->pad_.cpu_data();
const int* dilation_data = this->dilation_.cpu_data();
this->output_shape_.clear();
for (int i = 0; i < this->num_spatial_axes_; ++i) {
// i + 1 to skip channel axis
const int input_dim = this->input_shape(i + 1);
const int kernel_extent = dilation_data[i] * (kernel_shape_data[i] - 1) + 1;
const int output_dim = (input_dim + 2 * pad_data[i] - kernel_extent)
/ stride_data[i] + 1; //计算输出的数据的尺寸
this->output_shape_.push_back(output_dim);
}
}
/*遍历计算每个blob里面每张图像的卷积*/
template <typename Dtype>
void ConvolutionLayer<Dtype>::Forward_cpu(const vector<Blob<Dtype>*>& bottom,
const vector<Blob<Dtype>*>& top) {
const Dtype* weight = this->blobs_[0]->cpu_data();
for (int i = 0; i < bottom.size(); ++i) {/*遍历每个blob*/
const Dtype* bottom_data = bottom[i]->cpu_data();
Dtype* top_data = top[i]->mutable_cpu_data();
for (int n = 0; n < this->num_; ++n) { /*遍历每张图像*/
this->forward_cpu_gemm(bottom_data + n * this->bottom_dim_, weight,
top_data + n * this->top_dim_); /*见base_conv_layer.cpp中函数的实现*/
if (this->bias_term_) {
const Dtype* bias = this->blobs_[1]->cpu_data();
this->forward_cpu_bias(top_data + n * this->top_dim_, bias);/*如果有偏置，则要加上bias*/
}
}
}
}
//反向传播权重梯度与bias梯度、数据梯度
template <typename Dtype>
void ConvolutionLayer<Dtype>::Backward_cpu(const vector<Blob<Dtype>*>& top,
const vector<bool>& propagate_down, const vector<Blob<Dtype>*>& bottom) {
const Dtype* weight = this->blobs_[0]->cpu_data();
Dtype* weight_diff = this->blobs_[0]->mutable_cpu_diff();
for (int i = 0; i < top.size(); ++i) {
const Dtype* top_diff = top[i]->cpu_diff();
const Dtype* bottom_data = bottom[i]->cpu_data();
Dtype* bottom_diff = bottom[i]->mutable_cpu_diff();
// Bias gradient, if necessary.
// 如果有bias项，计算Bias导数
if (this->bias_term_ && this->param_propagate_down_[1]) {
Dtype* bias_diff = this->blobs_[1]->mutable_cpu_diff();
for (int n = 0; n < this->num_; ++n) {
this->backward_cpu_bias(bias_diff, top_diff + n * this->top_dim_);
}
}
if (this->param_propagate_down_[0] || propagate_down[i]) {
for (int n = 0; n < this->num_; ++n) {
// gradient w.r.t. weight. Note that we will accumulate diffs.
// 计算weights权重的梯度
if (this->param_propagate_down_[0]) {
this->weight_cpu_gemm(bottom_data + n * this->bottom_dim_,
top_diff + n * this->top_dim_, weight_diff);
}
// gradient w.r.t. bottom data, if necessary.
//计算botttom数据的梯度，向后传递
if (propagate_down[i]) {
this->backward_cpu_gemm(top_diff + n * this->top_dim_, weight,
bottom_diff + n * this->bottom_dim_);
}
}
}
}
}
#ifdef CPU_ONLY
STUB_GPU(ConvolutionLayer);
#endif
INSTANTIATE_CLASS(ConvolutionLayer);
} // namespace caffe
感谢：：：：：：http://blog.csdn.net/xizero00/article/details/51049858
扫一扫
1777
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
