max pooling in caffe

我们来看max pooling 在caffe 中怎么实现的吧

reshape

首先 reshap的时候：

  // If max pooling, we will initialize the vector index part.
  if (this->layer_param_.pooling_param().pool() ==
      PoolingParameter_PoolMethod_MAX && top.size() == 1) {
    max_idx_.Reshape(bottom[0]->num(), channels_, pooled_height_,
        pooled_width_);
  }

如是max pooling 则需要reshape max_idx 用来记录每次max pooling是 提取哪个地方的位置。
大小为$num\times channel \times pooled\_height \times pooled\_width$

forward

再看forward：

case PoolingParameter_PoolMethod_MAX:
    // Initialize 如果top有两个分支，就有top_mask 没研究这个。遇到再说，目前是进else分支
    if (use_top_mask) {
      top_mask = top[1]->mutable_cpu_data();
      caffe_set(top_count, Dtype(-1), top_mask);
    } else {
    //get 到 max_idx_的指针
      mask = max_idx_.mutable_cpu_data();
      caffe_set(top_count, -1, mask);
    }
    //top_data 全部变成大浮点数的相反数。方便后面的取max运算
    caffe_set(top_count, Dtype(-FLT_MAX), top_data);
    // The main loop 找最大值
    for (int n = 0; n < bottom[0]->num(); ++n) {
      for (int c = 0; c < channels_; ++c) {
        for (int ph = 0; ph < pooled_height_; ++ph) {
          for (int pw = 0; pw < pooled_width_; ++pw) {
            int hstart = ph * stride_h_ - pad_h_;
            int wstart = pw * stride_w_ - pad_w_;
            int hend = min(hstart + kernel_h_, height_);
            int wend = min(wstart + kernel_w_, width_);
            hstart = max(hstart, 0);
            wstart = max(wstart, 0);
            const int pool_index = ph * pooled_width_ + pw;
            for (int h = hstart; h < hend; ++h) {
              for (int w = wstart; w < wend; ++w) {
                const int index = h * width_ + w;
                if (bottom_data[index] > top_data[pool_index]) {
                  top_data[pool_index] = bottom_data[index];
                  if (use_top_mask) {
                    top_mask[pool_index] = static_cast<Dtype>(index);
                  } else {
                    mask[pool_index] = index;
                  }
                }
              }
            }
          }
        }
        // compute offset 移动指针位置
        bottom_data += bottom[0]->offset(0, 1);
        top_data += top[0]->offset(0, 1);
        if (use_top_mask) {
          top_mask += top[0]->offset(0, 1);
        } else {
          mask += top[0]->offset(0, 1);
        }
      }
    }
    break;

其中offset函数是这样定义的：

  inline int offset(const int n, const int c = 0, const int h = 0,
      const int w = 0) const {
    CHECK_GE(n, 0);
    CHECK_LE(n, num());
    CHECK_GE(channels(), 0);
    CHECK_LE(c, channels());
    CHECK_GE(height(), 0);
    CHECK_LE(h, height());
    CHECK_GE(width(), 0);
    CHECK_LE(w, width());
    return ((n * channels() + c) * height() + h) * width() + w;
  }

带入的都是0，1 也就是 平移$height \ times width$大小

backward

case PoolingParameter_PoolMethod_MAX:
    // The main loop
    if (use_top_mask) {
      top_mask = top[1]->cpu_data();
    } else {
      mask = max_idx_.cpu_data();
    }
    for (int n = 0; n < top[0]->num(); ++n) {
      for (int c = 0; c < channels_; ++c) {
        for (int ph = 0; ph < pooled_height_; ++ph) {
          for (int pw = 0; pw < pooled_width_; ++pw) {
            const int index = ph * pooled_width_ + pw;
            //找到对应位置 把上层的梯度加上去就好了
            const int bottom_index =
                use_top_mask ? top_mask[index] : mask[index];
            bottom_diff[bottom_index] += top_diff[index];
          }
        }
        bottom_diff += bottom[0]->offset(0, 1);
        top_diff += top[0]->offset(0, 1);
        if (use_top_mask) {
          top_mask += top[0]->offset(0, 1);
        } else {
          mask += top[0]->offset(0, 1);
        }
      }
    }
    break;