name: "VGG_ILSVRC_19_layer"

layer {
  name: "data"
  type: "ImageData"
  top: "data"
  top: "label"
  include {
    phase: TRAIN
  }
 
  image_data_param {
    batch_size: 12
    source: "../../fine_tuning_data/HAT_fineTuning_data/train_data_fineTuning.txt"
    root_folder: "../../fine_tuning_data/HAT_fineTuning_data/train_data/"
  }
} layer {
  name: "data"
  type: "ImageData"
  top: "data"
  top: "label"
  include {
    phase: TEST
  }
  transform_param {
    mirror: false
  }
  image_data_param {
    batch_size: 10
    source: "../../fine_tuning_data/HAT_fineTuning_data/test_data_fineTuning.txt"
    root_folder: "../../fine_tuning_data/HAT_fineTuning_data/test_data/"
  }
} layer {
  bottom:"data"
  top:"conv1_1"
  name:"conv1_1"
  type:"Convolution"
  convolution_param {
    num_output:64
    pad:1
    kernel_size:3
  }
}
layer {
  bottom:"conv1_1"
  top:"conv1_1"
  name:"relu1_1"
  type:"ReLU"
}
layer {
  bottom:"conv1_1"
  top:"conv1_2"
  name:"conv1_2"
  type:"Convolution"
  convolution_param {
    num_output:64
    pad:1
    kernel_size:3
  }
}
layer {
  bottom:"conv1_2"
  top:"conv1_2"
  name:"relu1_2"
  type:"ReLU"
}
layer {
  bottom:"conv1_2"
  top:"pool1"
  name:"pool1"
  type:"Pooling"
  pooling_param {
    pool:MAX
    kernel_size:2
    stride:2
  }
}
layer {
  bottom:"pool1"
  top:"conv2_1"
  name:"conv2_1"
  type:"Convolution"
  convolution_param {
    num_output:128
    pad:1
    kernel_size:3
  }
}
layer {
  bottom:"conv2_1"
  top:"conv2_1"
  name:"relu2_1"
  type:"ReLU"
}
layer {
  bottom:"conv2_1"
  top:"conv2_2"
  name:"conv2_2"
  type:"Convolution"
  convolution_param {
    num_output:128
    pad:1
    kernel_size:3
  }
}
layer {
  bottom:"conv2_2"
  top:"conv2_2"
  name:"relu2_2"
  type:"ReLU"
}
layer {
  bottom:"conv2_2"
  top:"pool2"
  name:"pool2"
  type:"Pooling"
  pooling_param {
    pool:MAX
    kernel_size:2
    stride:2
  }
}
layer {
  bottom:"pool2"
  top:"conv3_1"
  name: "conv3_1"
  type:"Convolution"
  convolution_param {
    num_output:256
    pad:1
    kernel_size:3
  }
}
layer {
  bottom:"conv3_1"
  top:"conv3_1"
  name:"relu3_1"
  type:"ReLU"
}
layer {
  bottom:"conv3_1"
  top:"conv3_2"
  name:"conv3_2"
  type:"Convolution"
  convolution_param {
    num_output:256
    pad:1
    kernel_size:3
  }
}
layer {
  bottom:"conv3_2"
  top:"conv3_2"
  name:"relu3_2"
  type:"ReLU"
}
layer {
  bottom:"conv3_2"
  top:"conv3_3"
  name:"conv3_3"
  type:"Convolution"
  convolution_param {
    num_output:256
    pad:1
    kernel_size:3
  }
}
layer {
  bottom:"conv3_3"
  top:"conv3_3"
  name:"relu3_3"
  type:"ReLU"
}
layer {
  bottom:"conv3_3"
  top:"conv3_4"
  name:"conv3_4"
  type:"Convolution"
  convolution_param {
    num_output:256
    pad:1
    kernel_size:3
  }
}
layer {
  bottom:"conv3_4"
  top:"conv3_4"
  name:"relu3_4"
  type:"ReLU"
}
layer {
  bottom:"conv3_4"
  top:"pool3"
  name:"pool3"
  type:"Pooling"
  pooling_param {
    pool:MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  bottom:"pool3"
  top:"conv4_1"
  name:"conv4_1"
  type:"Convolution"
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
  }
}
layer {
  bottom:"conv4_1"
  top:"conv4_1"
  name:"relu4_1"
  type:"ReLU"
}
layer {
  bottom:"conv4_1"
  top:"conv4_2"
  name:"conv4_2"
  type:"Convolution"
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
  }
}
layer {
  bottom:"conv4_2"
  top:"conv4_2"
  name:"relu4_2"
  type:"ReLU"
}
layer {
  bottom:"conv4_2"
  top:"conv4_3"
  name:"conv4_3"
  type:"Convolution"
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
  }
}
layer {
  bottom:"conv4_3"
  top:"conv4_3"
  name:"relu4_3"
  type:"ReLU"
}
layer {
  bottom:"conv4_3"
  top:"conv4_4"
  name:"conv4_4"
  type:"Convolution"
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
  }
}
layer {
  bottom:"conv4_4"
  top:"conv4_4"
  name:"relu4_4"
  type:"ReLU"
}
layer {
  bottom:"conv4_4"
  top:"pool4"
  name:"pool4"
  type:"Pooling"
  pooling_param {
    pool:MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  bottom:"pool4"
  top:"conv5_1"
  name:"conv5_1"
  type:"Convolution"
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
  }
}
layer {
  bottom:"conv5_1"
  top:"conv5_1"
  name:"relu5_1"
  type:"ReLU"
}
layer {
  bottom:"conv5_1"
  top:"conv5_2"
  name:"conv5_2"
  type:"Convolution"
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
  }
}
layer {
  bottom:"conv5_2"
  top:"conv5_2"
  name:"relu5_2"
  type:"ReLU"
}
layer {
  bottom:"conv5_2"
  top:"conv5_3"
  name:"conv5_3"
  type:"Convolution"
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
  }
}
layer {
  bottom:"conv5_3"
  top:"conv5_3"
  name:"relu5_3"
  type:"ReLU"
}
layer {
  bottom:"conv5_3"
  top:"conv5_4"
  name:"conv5_4"
  type:"Convolution"
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
  }
}
layer {
  bottom:"conv5_4"
  top:"conv5_4"
  name:"relu5_4"
  type:"ReLU"
}
layer {
  bottom:"conv5_4"
  top:"pool5"
  name:"pool5"
  type:"Pooling"
  pooling_param {
    pool:MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  bottom:"pool5"
  top:"fc6_"
  name:"fc6_"
  type:"InnerProduct"
  inner_product_param {
    num_output: 4096
  }
}
layer {
  bottom:"fc6_"
  top:"fc6_"
  name:"relu6"
  type:"ReLU"
}
layer {
  bottom:"fc6_"
  top:"fc6_"
  name:"drop6"
  type:"Dropout"
  dropout_param {
    dropout_ratio: 0.5
  }
}
layer {
  bottom:"fc6_"
  top:"fc7"
  name:"fc7"
  type:"InnerProduct"
  inner_product_param {
    num_output: 4096
  }
}
layer {
  bottom:"fc7"
  top:"fc7"
  name:"relu7"
  type:"ReLU"
}
layer {
  bottom:"fc7"
  top:"fc7"
  name:"drop7"
  type:"Dropout"
  dropout_param {
    dropout_ratio: 0.5
  }
}
layer {
  bottom:"fc7"
  top:"fc8_"
  name:"fc8_"
  type:"InnerProduct"
  inner_product_param {
    num_output: 27
  }
} layer {
  name: "sigmoid"
  type: "Sigmoid"
  bottom: "fc8_"
  top: "fc8_"
}  layer {
   name: "accuracy"
   type: "Accuracy"
   bottom: "fc8_"
   bottom: "label"
   top: "accuracy"
   include {
     phase: TEST
   }
 } layer {
  name: "loss"
  type: "EuclideanLoss"
  bottom: "fc8_"
  bottom: "label"
  top: "loss"
}

VGG_19 train_vali.prototxt file的更多相关文章

  1. 如何才能将Faster R-CNN训练起来?

    如何才能将Faster R-CNN训练起来? 首先进入 Faster RCNN 的官网啦,即:https://github.com/rbgirshick/py-faster-rcnn#installa ...

  2. SSD框架训练自己的数据集

    SSD demo中详细介绍了如何在VOC数据集上使用SSD进行物体检测的训练和验证.本文介绍如何使用SSD实现对自己数据集的训练和验证过程,内容包括: 1 数据集的标注2 数据集的转换3 使用SSD如 ...

  3. Faster-RCNN 训练自己的数据

    在前一篇随笔中,数据制作成了VOC2007格式,可以用于Faster-RCNN的训练. 1.针对数据的修改 修改datasets\VOCdevkit2007\VOCcode\VOCinit.m,我只做 ...

  4. caffe drawnet.py 用Python画网络框架

    在caffe中可以使用draw_net.py轻松地绘制卷积神经网络(CNN,Convolutional Neural Networks)的架构图.这个工具对于我们理解.学习甚至查错都有很大的帮助. 1 ...

  5. caffe实际运行中遇到的问题

    https://blog.csdn.net/u010417185/article/details/52649178 1.均值计算是否需要统一图像的尺寸? 在图像计算均值时,应该先统一图像的尺寸,否则会 ...

  6. [OpenCV] Install OpenCV 3.3 with DNN

    OpenCV 3.3 Aug 3, 2017 OpenCV 3.3 has been released with greatly improved Deep Learning module and l ...

  7. [PyImageSearch] Ubuntu16.04 使用深度学习和OpenCV实现物体检测

    上一篇博文中讲到如何用OpenCV实现物体分类,但是接下来这篇博文将会告诉你图片中物体的位置具体在哪里. 我们将会知道如何使用OpenCV‘s的dnn模块去加载一个预训练的物体检测网络,它能使得我们将 ...

  8. 【PyImageSearch】Ubuntu16.04使用OpenCV3.3.0实现图像分类

    这篇博文将会展示如何采用一个预训练的深度学习网络(模型)在ImageNet的数据集并把它当作输入图像. 首先说明,运行环境为Ubuntu16.04(或者MacOS),windows暂不支持,已经编译好 ...

  9. 机器学习进阶-目标追踪-SSD多进程执行 1.cv2.dnn.readnetFromCaffe(用于读取已经训练好的caffe模型) 2.delib.correlation_tracker(生成追踪器) 5.cv2.writer(将图片写入视频中) 6.cv2.dnn.blobFromImage(图片归一化) 10.multiprocessing.process(生成进程)

    1. cv2.dnn.readNetFromCaffe(prototxt, model)  用于进行SSD网络的caffe框架的加载 参数说明:prototxt表示caffe网络的结构文本,model ...

随机推荐

  1. 2016 - 1 -17 GCD学习总结

    一:GCD中的两个核心概念,队列与任务: 1.任务:执行什么操作.(代码块 block) 任务执行的类型分为以下两种: 1.1同步执行任务:在当前线程执行任务.不会开辟新的线程. 1.2异步执行任务: ...

  2. ios上 更改 状态栏(UIStatusBar)

    摘要 ios上 更改状态栏(UIStatusBar)的颜色 ios UIStatusBar statusBar 状态栏 更改状态栏颜色 目录[-] IOS上 关于状态栏的相关设置(UIStatusBa ...

  3. LeetCode----Word Ladder 2

    Given two words (start and end), and a dictionary, find all shortest transformation sequence(s) from ...

  4. swift系统学习控件篇:UITableView+UICollectionView

    工作之余,学习下swift大法.把自己的学习过程分享一下.当中的布局很乱,就表在意这些细节了.直接上代码: UITableView: // // ViewController.swift // UIt ...

  5. Could not launch "app_name"

    真机测试 不报错 编译通过后 Xcode总出这个错 process launch faild:NotFound-------解决办法 :重启设备

  6. 4、网页制作Dreamweaver(样式表CSS)

    样式表style 制作一个风格统一的网页,需要样式表对颜色.字体等属性的规范,同时也省去在body中多次定义的麻烦,所以一个样式表是必不可少的. 样式表有两种引用的方法:一种是直接写在html的< ...

  7. SqlSever2005 一千万条以上记录分页数据库优化经验总结

    http://www.cnblogs.com/jirigala/archive/2010/11/03/1868011.html 待测试???

  8. 获得供应商最近一次报价:OVER(PARTITION BY)函数用法的实际用法

    利用rownumber ,关键字partition进行小范围分页 方法一: --所有供应商对该产品最近的一次报价with oa as(select a.SupplierId ,UnitPrice,Pr ...

  9. tracking 问题解决

    1.dir,或者C++函数读文件名,不推荐.搞乱了名字 2. matio读写矩阵

  10. DNS劫持 DNS污染

    编号:1021时间:2016年6月24日17:23:50功能:DNS劫持 DNS污染URL:http://www.itechzero.com/dns-hijacking-dns-pollution-i ...