Caffe源码(七):ReLU,Sigmoid and Tanh

目录

简单介绍

ReLU 激活函数:

ReLu使得网络可以自行引入稀疏性,在没做预训练情况下,以ReLu为激活的网络性能优于其它激活函数。
数学表达式: y=max(0,x)

Sigmoid 激活函数:

sigmoid 激活函数在神经网络学习方面,可以将重点特征推向中央区,将非重点特征推向两侧区。
数学表达式: y=(1+exp(x))1

Tanh 激活函数:

Tanh 激活函数使得输出与输入的关系能保持非线性单调上升和下降关系,比sigmoid 函数延迟了饱和期,对神经网路的容错性好。
数学表达式: y=exp(x)exp(x)exp(x)+exp(x)

ReLU 主要函数

Forward_cpu 函数:

template <typename Dtype>
void ReLULayer<Dtype>::Forward_cpu(const vector<Blob<Dtype>*>& bottom,
    const vector<Blob<Dtype>*>& top) {
  const Dtype* bottom_data = bottom[0]->cpu_data();
  Dtype* top_data = top[0]->mutable_cpu_data();
  const int count = bottom[0]->count();
  Dtype negative_slope = this->layer_param_.relu_param().negative_slope(); //输入小于0时的斜率,默认为0;
  for (int i = 0; i < count; ++i) {
    top_data[i] = std::max(bottom_data[i], Dtype(0))
        + negative_slope * std::min(bottom_data[i], Dtype(0));
  }//输入大于零斜率为1,小于0斜率为negative_slope。
}

Backward_cpu 函数:

template <typename Dtype>
void ReLULayer<Dtype>::Backward_cpu(const vector<Blob<Dtype>*>& top,
    const vector<bool>& propagate_down,
    const vector<Blob<Dtype>*>& bottom) {
  if (propagate_down[0]) {
    const Dtype* bottom_data = bottom[0]->cpu_data();
    const Dtype* top_diff = top[0]->cpu_diff();
    Dtype* bottom_diff = bottom[0]->mutable_cpu_diff();
    const int count = bottom[0]->count();
    Dtype negative_slope = this->layer_param_.relu_param().negative_slope();
    for (int i = 0; i < count; ++i) {
      bottom_diff[i] = top_diff[i] * ((bottom_data[i] > 0)
          + negative_slope * (bottom_data[i] <= 0));
    }
  }
}

Sigmoid主要函数

Forward_cpu 函数:

template <typename Dtype>
void SigmoidLayer<Dtype>::Forward_cpu(const vector<Blob<Dtype>*>& bottom,
    const vector<Blob<Dtype>*>& top) {
  const Dtype* bottom_data = bottom[0]->cpu_data();
  Dtype* top_data = top[0]->mutable_cpu_data();
  const int count = bottom[0]->count();
  for (int i = 0; i < count; ++i) {
    top_data[i] = sigmoid(bottom_data[i]);
  }
}

sigmoid 函数定义如下:

template <typename Dtype>
inline Dtype sigmoid(Dtype x) {
  return 1. / (1. + exp(-x));
}

Backward_cpu 函数:

求导:

dydx=1(1+exp(x))2×(exp(x))=11+exp(x)×(111+exp(x))

template <typename Dtype>
void SigmoidLayer<Dtype>::Backward_cpu(const vector<Blob<Dtype>*>& top,
    const vector<bool>& propagate_down,
    const vector<Blob<Dtype>*>& bottom) {
  if (propagate_down[0]) {
    const Dtype* top_data = top[0]->cpu_data();
    const Dtype* top_diff = top[0]->cpu_diff();
    Dtype* bottom_diff = bottom[0]->mutable_cpu_diff();
    const int count = bottom[0]->count();
    for (int i = 0; i < count; ++i) {
      const Dtype sigmoid_x = top_data[i];
      bottom_diff[i] = top_diff[i] * sigmoid_x * (1. - sigmoid_x);
    }
  }
}

Tanh主要函数

Forward_cpu 函数:

template <typename Dtype>
void TanHLayer<Dtype>::Forward_cpu(const vector<Blob<Dtype>*>& bottom,
    const vector<Blob<Dtype>*>& top) {
  const Dtype* bottom_data = bottom[0]->cpu_data();
  Dtype* top_data = top[0]->mutable_cpu_data();
  const int count = bottom[0]->count();
  for (int i = 0; i < count; ++i) {
    top_data[i] = tanh(bottom_data[i]);
  }
}

Backward_cpu 函数:

求导:

dydx=(exp(x)+exp(x))2(exp(x)exp(x))2(exp(x)+exp(x))2=1(exp(x)exp(x)exp(x)+exp(x))2

template <typename Dtype>
void TanHLayer<Dtype>::Backward_cpu(const vector<Blob<Dtype>*>& top,
    const vector<bool>& propagate_down,
    const vector<Blob<Dtype>*>& bottom) {
  if (propagate_down[0]) {
    const Dtype* top_data = top[0]->cpu_data();
    const Dtype* top_diff = top[0]->cpu_diff();
    Dtype* bottom_diff = bottom[0]->mutable_cpu_diff();
    const int count = bottom[0]->count();
    Dtype tanhx;
    for (int i = 0; i < count; ++i) {
      tanhx = top_data[i];
      bottom_diff[i] = top_diff[i] * (1 - tanhx * tanhx);
    }
  }
}

猜你喜欢

转载自blog.csdn.net/seven_first/article/details/47703481