net_mel_2lang_bn.prototxt

name: "Berlin_net_mel"
layer {
  name: "data"
  type: "Data"
  top: "data"
  top: "label"
  include {
    phase: TRAIN
  }
  transform_param {
   mirror: false
  }
  data_param {
    source: "/home/mpss2015m_1/DeepAudio/iLID-Data/output/2016-01-26_melfilter/lmdb/train_en_de"
    batch_size: 64
    backend: LEVELDB
  }
}
layer {
  name: "data"
  type: "Data"
  top: "data"
  top: "label"
  include {
    phase: TEST
  }
  transform_param {
   mirror: false
  }
  data_param {
    source: "/home/mpss2015m_1/DeepAudio/iLID-Data/output/2016-01-26_melfilter/lmdb/test_en_de"
    batch_size: 64
    backend: LEVELDB
  }
}
layer {
  bottom: "data"
  top: "conv1"
  name: "conv1"
  type: "Convolution"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 12
    kernel_size: 6
    stride: 1
    weight_filler {
      type: "gaussian" # initialize the filters from a Gaussian
      std: 0.01        # distribution with stdev 0.01 (default mean: 0)
    }
    bias_filler {
      type: "constant" # initialize the biases to zero (0)
        value: 0
    }
  }
}
layer {
  bottom: "conv1"
  top: "conv1"
  name: "Relu1"
  type: "ReLU"
}
layer {
  name: "bn1"
  type: "BatchNorm"
  bottom: "conv1"
  top: "bn1"
  param {
    lr_mult: 0
  }
  param {
    lr_mult: 0
  }
  param {
    lr_mult: 0
  }
}
layer {
  bottom: "bn1"
  top: "pool1"
  name: "pool1"
  type: "Pooling"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  bottom: "pool1"
  top: "conv2"
  name: "conv2"
  type: "Convolution"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 12
    pad: 1
    kernel_size: 6
    stride: 1
    weight_filler {
      type: "gaussian" # initialize the filters from a Gaussian
      std: 0.01        # distribution with stdev 0.01 (default mean: 0)
    }
    bias_filler {
      type: "constant" # initialize the biases to zero (0)
        value: 0
    }
  }
}
layer {
  bottom: "conv2"
  top: "conv2"
  name: "relu2"
  type: "ReLU"
}
layer {
  name: "bn2"
  type: "BatchNorm"
  bottom: "conv2"
  top: "bn2"
  param {
    lr_mult: 0
  }
  param {
    lr_mult: 0
  }
  param {
    lr_mult: 0
  }
}
layer {
  bottom: "bn2"
  top: "pool2"
  name: "pool2"
  type: "Pooling"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  bottom: "pool2"
  top: "conv3"
  name: "conv3"
  type: "Convolution"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 12
    pad: 1
    kernel_size: 6
    weight_filler {
      type: "gaussian" # initialize the filters from a Gaussian
      std: 0.01        # distribution with stdev 0.01 (default mean: 0)
    }
    bias_filler {
      type: "constant" # initialize the biases to zero (0)
        value: 0
    }
  }
}
layer {
  bottom: "conv3"
  top: "conv3"
  name: "relu3"
  type: "ReLU"
}
layer {
  name: "bn3"
  type: "BatchNorm"
  bottom: "conv3"
  top: "bn3"
  param {
    lr_mult: 0
  }
  param {
    lr_mult: 0
  }
  param {
    lr_mult: 0
  }
}
layer {
  bottom: "bn3"
  top: "pool3"
  name: "pool3"
  type: "Pooling"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  bottom: "pool3"
  top: "fc4"
  name: "fc4"
  type: "InnerProduct"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  inner_product_param {
    num_output: 1024
    weight_filler {
      type: "gaussian" # initialize the filters from a Gaussian
      std: 0.01        # distribution with stdev 0.01 (default mean: 0)
    }
    bias_filler {
      type: "constant" # initialize the biases to zero (0)
        value: 0
    }
  }
}
layer {
  bottom: "fc4"
  top: "fc4"
  name: "relu4"
  type: "ReLU"
}
layer {
  name: "fc5"
  type: "InnerProduct"
  bottom: "fc4"
  top: "fc5"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  inner_product_param {
    num_output: 2
    weight_filler {
      type: "gaussian" # initialize the filters from a Gaussian
      std: 0.01        # distribution with stdev 0.01 (default mean: 0)
    }
    bias_filler {
      type: "constant" # initialize the biases to zero (0)
        value: 0
    }
  }
}
layer {
  name: "accuracy"
  type: "Accuracy"
  bottom: "fc5"
  bottom: "label"
  top: "accuracy"
  include {
    phase: TEST
  }
}
layer {
  name: "loss"
  type: "SoftmaxWithLoss"
  bottom: "fc5"
  bottom: "label"
  top: "loss"
}