原始caffe添加CenterLoss_layer

2023-10-29 03:08
文章标签 caffe 原始 layer centerloss

本文主要是介绍原始caffe添加CenterLoss_layer，希望对大家解决编程问题提供一定的参考价值，需要的开发者们随着小编来一起学习吧！

CenterLoss_layer可以在原分类的基础上（某种程度上）可提升几个点左右（博主测试提升6个点），还不占前向时间，好东西啊！

原理这里不介绍了，网上应有尽有！这里只是介绍如何在caffe中如何添加CenterLoss_layer这样的新层！

 

第一步：修改caffe.proto以添加消息机制

1、在message LayerParameter{}中添加如下代码：

 

optional CenterLossParameter center_loss_param = 147;


2、在caffe.proto文末添加

 

 

message CenterLossParameter {optional uint32 num_output = 1; // The number of outputs for the layeroptional FillerParameter center_filler = 2; // The filler for the centers// The first axis to be lumped into a single inner product computation;// all preceding axes are retained in the output.// May be negative to index from the end (e.g., -1 for the last axis).optional int32 axis = 3 [default = 1];
}

 

 

 

 

 

第二步：添加头文件center_loss_layer.hpp

 

#ifndef CAFFE_CENTER_LOSS_LAYER_HPP_
#define CAFFE_CENTER_LOSS_LAYER_HPP_#include <vector>#include "caffe/blob.hpp"
#include "caffe/layer.hpp"
#include "caffe/proto/caffe.pb.h"#include "caffe/layers/loss_layer.hpp"namespace caffe {template <typename Dtype>
class CenterLossLayer : public LossLayer<Dtype> {public:explicit CenterLossLayer(const LayerParameter& param): LossLayer<Dtype>(param) {}virtual void LayerSetUp(const vector<Blob<Dtype>*>& bottom,const vector<Blob<Dtype>*>& top);virtual void Reshape(const vector<Blob<Dtype>*>& bottom,const vector<Blob<Dtype>*>& top);virtual inline const char* type() const { return "CenterLoss"; }virtual inline int ExactNumBottomBlobs() const { return 2; }virtual inline int ExactNumTopBlobs() const { return -1; }protected:virtual void Forward_cpu(const vector<Blob<Dtype>*>& bottom,const vector<Blob<Dtype>*>& top);virtual void Forward_gpu(const vector<Blob<Dtype>*>& bottom,const vector<Blob<Dtype>*>& top);virtual void Backward_cpu(const vector<Blob<Dtype>*>& top,const vector<bool>& propagate_down, const vector<Blob<Dtype>*>& bottom);virtual void Backward_gpu(const vector<Blob<Dtype>*>& top,const vector<bool>& propagate_down, const vector<Blob<Dtype>*>& bottom);int M_;int K_;int N_;Blob<Dtype> distance_;Blob<Dtype> variation_sum_;
};}  // namespace caffe#endif  // CAFFE_CENTER_LOSS_LAYER_HPP_

 

 

 

 

 

第三步：添加center_loss_layer.cpp以及center_loss_layer.cu

 

#include <vector>#include "caffe/filler.hpp"
#include "caffe/layers/center_loss_layer.hpp"
#include "caffe/util/math_functions.hpp"namespace caffe {template <typename Dtype>
void CenterLossLayer<Dtype>::LayerSetUp(const vector<Blob<Dtype>*>& bottom,const vector<Blob<Dtype>*>& top) {const int num_output = this->layer_param_.center_loss_param().num_output();  N_ = num_output;const int axis = bottom[0]->CanonicalAxisIndex(this->layer_param_.center_loss_param().axis());// Dimensions starting from "axis" are "flattened" into a single// length K_ vector. For example, if bottom[0]'s shape is (N, C, H, W),// and axis == 1, N inner products with dimension CHW are performed.K_ = bottom[0]->count(axis);// Check if we need to set up the weightsif (this->blobs_.size() > 0) {LOG(INFO) << "Skipping parameter initialization";} else {this->blobs_.resize(1);// Intialize the weightvector<int> center_shape(2);center_shape[0] = N_;center_shape[1] = K_;this->blobs_[0].reset(new Blob<Dtype>(center_shape));// fill the weightsshared_ptr<Filler<Dtype> > center_filler(GetFiller<Dtype>(this->layer_param_.center_loss_param().center_filler()));center_filler->Fill(this->blobs_[0].get());}  // parameter initializationthis->param_propagate_down_.resize(this->blobs_.size(), true);
}template <typename Dtype>
void CenterLossLayer<Dtype>::Reshape(const vector<Blob<Dtype>*>& bottom,const vector<Blob<Dtype>*>& top) {CHECK_EQ(bottom[1]->channels(), 1);CHECK_EQ(bottom[1]->height(), 1);CHECK_EQ(bottom[1]->width(), 1);M_ = bottom[0]->num();// The top shape will be the bottom shape with the flattened axes dropped,// and replaced by a single axis with dimension num_output (N_).LossLayer<Dtype>::Reshape(bottom, top);distance_.ReshapeLike(*bottom[0]);variation_sum_.ReshapeLike(*this->blobs_[0]);
}template <typename Dtype>
void CenterLossLayer<Dtype>::Forward_cpu(const vector<Blob<Dtype>*>& bottom,const vector<Blob<Dtype>*>& top) {const Dtype* bottom_data = bottom[0]->cpu_data();const Dtype* label = bottom[1]->cpu_data();const Dtype* center = this->blobs_[0]->cpu_data();Dtype* distance_data = distance_.mutable_cpu_data();// the i-th distance_datafor (int i = 0; i < M_; i++) {const int label_value = static_cast<int>(label[i]);// D(i,:) = X(i,:) - C(y(i),:)caffe_sub(K_, bottom_data + i * K_, center + label_value * K_, distance_data + i * K_);}Dtype dot = caffe_cpu_dot(M_ * K_, distance_.cpu_data(), distance_.cpu_data());Dtype loss = dot / M_ / Dtype(2);top[0]->mutable_cpu_data()[0] = loss;
}template <typename Dtype>
void CenterLossLayer<Dtype>::Backward_cpu(const vector<Blob<Dtype>*>& top,const vector<bool>& propagate_down,const vector<Blob<Dtype>*>& bottom) {// Gradient with respect to centersif (this->param_propagate_down_[0]) {const Dtype* label = bottom[1]->cpu_data();Dtype* center_diff = this->blobs_[0]->mutable_cpu_diff();Dtype* variation_sum_data = variation_sum_.mutable_cpu_data();const Dtype* distance_data = distance_.cpu_data();// \sum_{y_i==j}caffe_set(N_ * K_, (Dtype)0., variation_sum_.mutable_cpu_data());for (int n = 0; n < N_; n++) {int count = 0;for (int m = 0; m < M_; m++) {const int label_value = static_cast<int>(label[m]);if (label_value == n) {count++;caffe_sub(K_, variation_sum_data + n * K_, distance_data + m * K_, variation_sum_data + n * K_);}}caffe_axpy(K_, (Dtype)1./(count + (Dtype)1.), variation_sum_data + n * K_, center_diff + n * K_);}}// Gradient with respect to bottom data if (propagate_down[0]) {caffe_copy(M_ * K_, distance_.cpu_data(), bottom[0]->mutable_cpu_diff());caffe_scal(M_ * K_, top[0]->cpu_diff()[0] / M_, bottom[0]->mutable_cpu_diff());}if (propagate_down[1]) {LOG(FATAL) << this->type()<< " Layer cannot backpropagate to label inputs.";}
}#ifdef CPU_ONLY
STUB_GPU(CenterLossLayer);
#endifINSTANTIATE_CLASS(CenterLossLayer);
REGISTER_LAYER_CLASS(CenterLoss);}  // namespace caffe

 

#include <vector>#include "caffe/filler.hpp"
#include "caffe/layers/center_loss_layer.hpp"
#include "caffe/util/math_functions.hpp"namespace caffe {template <typename Dtype>
__global__ void Compute_distance_data_gpu(int nthreads, const int K, const Dtype* bottom,const Dtype* label, const Dtype* center, Dtype* distance) {CUDA_KERNEL_LOOP(index, nthreads) {int m = index / K;int k = index % K;const int label_value = static_cast<int>(label[m]);// distance(i) = x(i) - c_{y(i)}distance[index] = bottom[index] - center[label_value * K + k];}
}template <typename Dtype>
__global__ void Compute_center_diff_gpu(int nthreads, const int M, const int K, const Dtype* label, const Dtype* distance, Dtype* variation_sum, Dtype* center_diff) {CUDA_KERNEL_LOOP(index, nthreads) {int count = 0;for (int m = 0; m < M; m++) {const int label_value = static_cast<int>(label[m]);if (label_value == index) {count++;for (int k = 0; k < K; k++) {variation_sum[index * K + k] -= distance[m * K + k];}}}for (int k = 0; k < K; k++) {center_diff[index * K + k] = variation_sum[index * K + k] /(count + (Dtype)1.);}}
}template <typename Dtype>
void CenterLossLayer<Dtype>::Forward_gpu(const vector<Blob<Dtype>*>& bottom,const vector<Blob<Dtype>*>& top) {int nthreads = M_ * K_;Compute_distance_data_gpu<Dtype><<<CAFFE_GET_BLOCKS(nthreads),CAFFE_CUDA_NUM_THREADS>>>(nthreads, K_, bottom[0]->gpu_data(), bottom[1]->gpu_data(),this->blobs_[0]->gpu_data(), distance_.mutable_gpu_data());Dtype dot;caffe_gpu_dot(M_ * K_, distance_.gpu_data(), distance_.gpu_data(), &dot);Dtype loss = dot / M_ / Dtype(2);top[0]->mutable_cpu_data()[0] = loss;
}template <typename Dtype>
void CenterLossLayer<Dtype>::Backward_gpu(const vector<Blob<Dtype>*>& top,const vector<bool>& propagate_down,const vector<Blob<Dtype>*>& bottom) {int nthreads = N_;caffe_gpu_set(N_ * K_, (Dtype)0., variation_sum_.mutable_cpu_data());Compute_center_diff_gpu<Dtype><<<CAFFE_GET_BLOCKS(nthreads),CAFFE_CUDA_NUM_THREADS>>>(nthreads, M_, K_, bottom[1]->gpu_data(), distance_.gpu_data(), variation_sum_.mutable_cpu_data(), this->blobs_[0]->mutable_gpu_diff());if (propagate_down[0]) {caffe_gpu_scale(M_ * K_, top[0]->cpu_diff()[0] / M_, distance_.gpu_data(), bottom[0]->mutable_gpu_diff());}if (propagate_down[1]) {LOG(FATAL) << this->type()<< " Layer cannot backpropagate to label inputs.";}
}INSTANTIATE_LAYER_GPU_FUNCS(CenterLossLayer);}  // namespace caffe

 

 

 

 

 

第四步：重新编译caffe

 

任何问题请加唯一QQ2258205918（名称samylee）！

或唯一VX：samylee_csdn

 

这篇关于原始caffe添加CenterLoss_layer的文章就介绍到这儿，希望我们推荐的文章对编程师们有所帮助！


http://www.chinasem.cn/article/297852。

相关文章

[Linux Kernel Block Layer第一篇] block layer架构设计

[Linux Kernel Block Layer第一篇] block layer架构设计

目录 1. single queue架构 2. multi-queue架构（blk-mq)  3. 问题 随着SSD快速存储设备的发展，内核社区越发发现，存储的性能瓶颈从硬件存储设备转移到了内核block layer，主要因为当时的内核block layer是single hw queue的架构，导致cpu锁竞争问题严重，本文先提纲挈领的介绍内核block layer的架构演进，然
阅读更多...
使用openpose caffe源码框架训练车辆模型常见错误及解决办法

使用openpose caffe源码框架训练车辆模型常见错误及解决办法

错误1：what():  Error: mSources.size() != mProbabilities.size() at 51, OPDataLayer, src/caffe/openpose/layers/oPDataLayer.cpp 原因：这是因为在网络模型中数据源sources和probabilities设置的参数个数不一样导致的，一个数据源对应一个概率 解决方法：只需要将网络文
阅读更多...
caffe训练openpose相关资源

caffe训练openpose相关资源

CPMTransformationParameter参数解析： https://www.jianshu.com/p/063a2159f0f2 genLMDB.py： https://www.jianshu.com/p/1cae32cbd36d OpenPose 参数说明： https://blog.csdn.net/zziahgf/article/details/84668319 openp
阅读更多...
在Ubuntu16环境下用QT调试caffe

在Ubuntu16环境下用QT调试caffe

用qt调试caffe需要配置一些东西具体方法可以参考一下链接： https://www.zhihu.com/question/27987666/answer/80728225 https://blog.csdn.net/xg123321123/article/details/52817658
阅读更多...
ubuntu16.04 caffe python3.5 安装

ubuntu16.04 caffe python3.5 安装

在ubuntu16.04 下 caffe安装python3.5可以参考 https://blog.csdn.net/tingtie1438/article/details/82085199 但这里面有一些需要补充和修正其中的一些东西 一、其中“二. 安装依赖项”里面是这样的： sudo apt-get install python3-pipsudo apt-get install pytho
阅读更多...
解决caffe 编译过程中protobuf版本冲突的问题

解决caffe 编译过程中protobuf版本冲突的问题

在编译caffe python3版本时一直会出现如下错误，（安装caffe python3具体方法可参考：https://blog.csdn.net/tingtie1438/article/details/82085199 ）： 通过其错误信息可知是protobuf出了问题，现在网上教程一般都是默认安装的 libprotobuf-dev 和 protobuf-compiler，对于pytho
阅读更多...
ubuntu16.04 caffe(github源码cpu)+python3.5+opencv3.4.5安装编译

ubuntu16.04 caffe(github源码cpu)+python3.5+opencv3.4.5安装编译

https://www.cnblogs.com/hanjianjian90/p/10604926.html
阅读更多...
android xml之Drawable 篇 --------shape和selector和layer-list的

android xml之Drawable 篇 --------shape和selector和layer-list的

转自 ： http://blog.csdn.net/brokge/article/details/9713041 <shape>和<selector>在Android UI设计中经常用到。比如我们要自定义一个圆角Button，点击Button有些效果的变化，就要用到<shape>和<selector>。 可以这样说，<shape>和<selector>在美化控件中的作用是至关重要。 在
阅读更多...
Android 使用原始xml资源

Android 使用原始xml资源

原始XML资源一般保存在/res/xml/路径下。 XML文件中访问语法格式： @[<package_name>:]xml/file_name Java代码中访问语法格式： [<package_name>.]R.xml.<file_name> 在JAVA程序中获取实际的XML文档，可以通过Resources的如下两个方法来实现： XmlResourceParser getXml(int
阅读更多...
Layer Normalization论文解读

Layer Normalization论文解读

基本信息 作者JL Badoi发表时间2016期刊NIPS网址https://arxiv.org/abs/1607.06450v1 研究背景 1. What’s known 既往研究已证实 batch Normalization对属于同一个Batch中的数据长度要求是相同的，不适合处理序列型的数据。因此它在NLP领域的RNN上效果并不显著，但在CV领域的CNN上效果显著。 2. What’s
阅读更多...

Copyright China编程 23002807@qq.com

沪ICP备2023033072号-2