movepredict.prototxt

name: "LogReg"
input: "data"
input_dim: 1
input_dim: 2
input_dim: 19
input_dim: 19

#this part should be the same in learning and prediction network
layers {
  name: "conv1_7x7_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "data"
  top: "conv2"
  convolution_param {
    num_output: 128
    kernel_size: 7
    pad: 3
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu2"
  type: RELU
  bottom: "conv2"
  top: "conv2"
}

layers {
  name: "conv2_5x5_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv2"
  top: "conv3"
  convolution_param {
    num_output: 128
    kernel_size: 5
    pad: 2
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu3"
  type: RELU
  bottom: "conv3"
  top: "conv3"
}

layers {
  name: "conv3_5x5_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv3"
  top: "conv4"
  convolution_param {
    num_output: 128
    kernel_size: 5
    pad: 2
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu4"
  type: RELU
  bottom: "conv4"
  top: "conv4"
}

layers {
  name: "conv4_5x5_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv4"
  top: "conv5"
  convolution_param {
    num_output: 128
    kernel_size: 5
    pad: 2
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu5"
  type: RELU
  bottom: "conv5"
  top: "conv5"
}

layers {
  name: "conv5_5x5_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv5"
  top: "conv6"
  convolution_param {
    num_output: 128
    kernel_size: 5
    pad: 2
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu6"
  type: RELU
  bottom: "conv6"
  top: "conv6"
}


layers {
  name: "conv6_5x5_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv6"
  top: "conv7"
  convolution_param {
    num_output: 128
    kernel_size: 5
    pad: 2
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu7"
  type: RELU
  bottom: "conv7"
  top: "conv7"
}


layers {
  name: "conv7_5x5_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv7"
  top: "conv8"
  convolution_param {
    num_output: 128
    kernel_size: 5
    pad: 2
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu8"
  type: RELU
  bottom: "conv8"
  top: "conv8"
}


layers {
  name: "conv8_3x3_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv8"
  top: "conv9"
  convolution_param {
    num_output: 128
    kernel_size: 3
    pad: 1
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu9"
  type: RELU
  bottom: "conv9"
  top: "conv9"
}

layers {
  name: "conv9_3x3_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv9"
  top: "conv10"
  convolution_param {
    num_output: 128
    kernel_size: 3
    pad: 1
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu10"
  type: RELU
  bottom: "conv10"
  top: "conv10"
}

layers {
  name: "conv10_3x3_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv10"
  top: "conv11"
  convolution_param {
    num_output: 128
    kernel_size: 3
    pad: 1
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu11"
  type: RELU
  bottom: "conv11"
  top: "conv11"
}

layers {
  name: "conv11_3x3_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv11"
  top: "conv12"
  convolution_param {
    num_output: 128
    kernel_size: 3
    pad: 1
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu12"
  type: RELU
  bottom: "conv12"
  top: "conv12"
}

layers {
  name: "conv12_3x3_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv12"
  top: "conv13"
  convolution_param {
    num_output: 128
    kernel_size: 3
    pad: 1
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu13"
  type: RELU
  bottom: "conv13"
  top: "conv13"
}

layers {
  name: "conv13_3x3_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv13"
  top: "conv14"
  convolution_param {
    num_output: 128
    kernel_size: 3
    pad: 1
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu14"
  type: RELU
  bottom: "conv14"
  top: "conv14"
}

layers {
  name: "conv14_3x3_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv14"
  top: "conv15"
  convolution_param {
    num_output: 128
    kernel_size: 3
    pad: 1
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu15"
  type: RELU
  bottom: "conv15"
  top: "conv15"
}

layers {
  name: "conv15_3x3_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv15"
  top: "conv16"
  convolution_param {
    num_output: 128
    kernel_size: 3
    pad: 1
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu16"
  type: RELU
  bottom: "conv16"
  top: "conv16"
}

layers {
  name: "conv16_3x3_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv16"
  top: "conv17"
  convolution_param {
    num_output: 128
    kernel_size: 3
    pad: 1
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu17"
  type: RELU
  bottom: "conv17"
  top: "conv17"
}

layers {
  name: "conv17_3x3_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv17"
  top: "conv18"
  convolution_param {
    num_output: 128
    kernel_size: 3
    pad: 1
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu18"
  type: RELU
  bottom: "conv18"
  top: "conv18"
}

layers {
  name: "conv18_3x3_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv18"
  top: "conv19"
  convolution_param {
    num_output: 128
    kernel_size: 3
    pad: 1
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu19"
  type: RELU
  bottom: "conv19"
  top: "conv19"
}

layers {
  name: "conv19_3x3_128"
  type: CONVOLUTION
  blobs_lr: 1.0
  blobs_lr: 2.0
  bottom: "conv19"
  top: "conv20"
  convolution_param {
    num_output: 128
    kernel_size: 3
    pad: 1
    weight_filler {
      type: "xavier"
      }
      bias_filler {
      type: "constant"
      }
    }
}

layers {
  name: "relu20"
  type: RELU
  bottom: "conv20"
  top: "conv20"
}



layers {
  name: "ip"
  type: INNER_PRODUCT
  bottom: "conv20"
  # I had to comment out that line to make it work
  #top: "ip_zw"
  inner_product_param {
    num_output: 361
    weight_filler {
      type: "xavier"
      }
    bias_filler {
      type: "constant"
      }
   }
}

#layers {
#  name: "flat"
#  type: FLATTEN
#  bottom: "conv8"
#  top: "ip_zw"
#}




#only prediction
layers {
  name: "softmax"
  type: SOFTMAX
  bottom: "ip_zw"
  top: "ip"
}