关于测试自己的单张图片所需要的lenet.prototxt（deploy）文件

最新推荐文章于 2018-09-17 08:50:25 发布

转载最新推荐文章于 2018-09-17 08:50:25 发布 · 1k 阅读

文章标签：

#caffe #自己数据 #Layer #dim #prototxt

caffe 同时被 2 个专栏收录

39 篇文章

订阅专栏

Caffe

32 篇文章

订阅专栏

本文介绍如何从训练测试配置文件cifar10_quick_train_test.prototxt转换为部署文件deploy.prototxt的过程。主要包括删除测试层信息、修改训练层数据源为输入层并调整最后一层为Softmax层。

Depoly文件其实就是在刚刚定义定义层的参数的那个文件基础上进行修改的，

delpoy文件只记录了网络结构，而并没有里面的反向计算，也不需要计算误差。

我使用cifar10_quick_train_test.ptototxt进行修改，修改成对应的deploy文件.

cifar10_quick_train_test.prototxt文件内容如下：

name: "CIFAR10_quick"
layer {
  name: "cifar"
  type: "Data"
  top: "data"
  top: "label"
  include {
    phase: TRAIN
  }
  transform_param {
    mean_file: "C:/Users/Administrator/Desktop/data12345/mean.binaryproto"
  }
  data_param {
    source: "C:/Users/Administrator/Desktop/data12345/mtrainldb"
    batch_size: 50
    backend: LEVELDB
  }
}
layer {
  name: "cifar"
  type: "Data"
  top: "data"
  top: "label"
  include {
    phase: TEST
  }
  transform_param {
    mean_file: "C:/Users/Administrator/Desktop/data12345/mean.binaryproto"
  }
  data_param {
    source: "C:/Users/Administrator/Desktop/data12345/mtestldb"
    batch_size: 2
    backend: LEVELDB
  }
}
layer {
  name: "conv1"
  type: "Convolution"
  bottom: "data"
  top: "conv1"
  param {
    lr_mult: 1
  }
  param {
    lr_mult: 2
  }
  convolution_param {
    num_output: 16
    pad: 2
    kernel_size: 5
    stride: 1
    weight_filler {
      type: "gaussian"
      std: 0.0001
    }
    bias_filler {
      type: "constant"
    }
  }
}
layer {
  name: "pool1"
  type: "Pooling"
  bottom: "conv1"
  top: "pool1"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "relu1"
  type: "ReLU"
  bottom: "pool1"
  top: "pool1"
}
layer {
  name: "conv2"
  type: "Convolution"
  bottom: "pool1"
  top: "conv2"
  param {
    lr_mult: 1
  }
  param {
    lr_mult: 2
  }
  convolution_param {
    num_output: 24
    pad: 2
    kernel_size: 5
    stride: 1
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
    }
  }
}
layer {
  name: "relu2"
  type: "ReLU"
  bottom: "conv2"
  top: "conv2"
}
layer {
  name: "pool2"
  type: "Pooling"
  bottom: "conv2"
  top: "pool2"
  pooling_param {
    pool: AVE
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "conv3"
  type: "Convolution"
  bottom: "pool2"
  top: "conv3"
  param {
    lr_mult: 1
  }
  param {
    lr_mult: 2
  }
  convolution_param {
    num_output: 50
    pad: 2
    kernel_size: 5
    stride: 1
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
    }
  }
}
layer {
  name: "relu3"
  type: "ReLU"
  bottom: "conv3"
  top: "conv3"
}
layer {
  name: "pool3"
  type: "Pooling"
  bottom: "conv3"
  top: "pool3"
  pooling_param {
    pool: AVE
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "ip1"
  type: "InnerProduct"
  bottom: "pool3"
  top: "ip1"
  param {
    lr_mult: 1
  }
  param {
    lr_mult: 2
  }
  inner_product_param {
    num_output: 786
    weight_filler {
      type: "gaussian"
      std: 0.1
    }
    bias_filler {
      type: "constant"
    }
  }
}
layer {
  name: "ip2"
  type: "InnerProduct"
  bottom: "ip1"
  top: "ip2"
  param {
    lr_mult: 1
  }
  param {
    lr_mult: 2
  }
  inner_product_param {
    num_output: 2
    weight_filler {
      type: "gaussian"
      std: 0.1
    }
    bias_filler {
      type: "constant"
    }
  }
}
layer {
  name: "accuracy"
  type: "Accuracy"
  bottom: "ip2"
  bottom: "label"
  top: "accuracy"
  include {
    phase: TEST
  }
}
layer {
  name: "loss"
  type: "SoftmaxWithLoss"
  bottom: "ip2"
  bottom: "label"
  top: "loss"
}

修改方法如下：

（1）删掉有关于TEST层的信息。共有两个地方

第一个是在开头处：

layer {
  name: "cifar"
  type: "Data"
  top: "data"
  top: "label"
  include {
    phase: TEST
  }
  transform_param {
    mean_file: "C:/Users/Administrator/Desktop/data12345/mean.binaryproto"
  }
  data_param {
    source: "C:/Users/Administrator/Desktop/data12345/mtestldb"
    batch_size: 2
    backend: LEVELDB
  }
}

还有一处是结尾处，accuracy

layer {
  name: "accuracy"
  type: "Accuracy"
  bottom: "ip2"
  bottom: "label"
  top: "accuracy"
  include {
    phase: TEST
  }
}

（2）将开头Train层的信息修改：

原来的信息如下：

layer {
  name: "cifar"
  type: "Data"
  top: "data"
  top: "label"
  include {
    phase: TRAIN
  }
  transform_param {
    mean_file: "C:/Users/Administrator/Desktop/data12345/mean.binaryproto"
  }
  data_param {
    source: "C:/Users/Administrator/Desktop/data12345/mtrainldb"
    batch_size: 50
    backend: LEVELDB
  }
}

修改后如下：

layer {
  name: "data"
  type: "Input"
  top: "data"
  input_param { shape: { dim: 64 dim: 1 dim: 28 dim: 28 } }
}

其中

dim: 64 dim: 3 dim: 28 dim: 28

第一个代表了batch_size，第二个代表颜色通道数，像我就是彩色图片，所以是3，黑白是1，后面俩数分别代表图片的width和height。

（3）修改最后一层

原来是这样子的：

layer {
  name: "loss"
  type: "SoftmaxWithLoss"
  bottom: "ip2"
  bottom: "label"
  top: "loss"
}

现在修改如下：

layer {
  name: "prob"
  type: "Softmax"
  bottom: "fc8"
  top: "prob"
}

以上部分参考了：http://blog.youkuaiyun.com/redfivehit/article/details/72902418