LLCNN.prototxt

name: "LLCNN-18"
layer {
  name: "data"
  type: "HDF5Data"
  top: "data"
  top: "label"
  hdf5_data_param {
    source: "examples/LLNet/1channel/train.txt"
    batch_size: 64
  }
  include: { phase: TRAIN }
}
layer {
  name: "data"
  type: "HDF5Data"
  top: "data"
  top: "label"
  hdf5_data_param {
    source: "examples/LLNet/1channel/test.txt"
    batch_size: 16
  }
  include: { phase: TEST }
}

layer {
    bottom: "data"
    top: "conv1"
    name: "conv1"
    type: "Convolution"
    convolution_param {
        num_output: 64
        kernel_size: 3
        pad: 1
        stride: 1
        weight_filler {
            type: "msra"
        }
        
    }
}

layer {
    bottom: "conv1"
    top: "conv1"
    name: "relu1"
    type: "ReLU"
}

layer {
    bottom: "conv1"
    top: "conv2"
    name: "conv2"
    type: "Convolution"
    convolution_param {
        num_output: 64
        kernel_size: 1
        pad: 0
        stride: 1
        weight_filler {
            type: "msra"
        }
        
    }
}

layer {
    bottom: "conv1"
    top: "conv3"
    name: "conv3"
    type: "Convolution"
    convolution_param {
        num_output: 64
        kernel_size: 3
        pad: 1
        stride: 1
        weight_filler {
            type: "msra"
        }
        
    }
}

layer {
    bottom: "conv3"
    top: "conv3"
    name: "relu3"
    type: "ReLU"
}

layer {
    bottom: "conv3"
    top: "conv4"
    name: "conv4"
    type: "Convolution"
    convolution_param {
        num_output: 64
        kernel_size: 3
        pad: 1
        stride: 1
        weight_filler {
            type: "msra"
        }
        
    }
}

layer {
    bottom: "conv2"
    bottom: "conv4"
    top: "res2a"
    name: "res2a"
    type: "Eltwise"
    eltwise_param {
        operation: SUM
    }
}

layer {
    bottom: "res2a"
    top: "res2a"
    name: "res2a_relu"
    type: "ReLU"
}

layer {
    bottom: "res2a"
    top: "conv5"
    name: "conv5"
    type: "Convolution"
    convolution_param {
        num_output: 64
        kernel_size: 3
        pad: 1
        stride: 1
        weight_filler {
            type: "msra"
        }
        
    }
}

layer {
    bottom: "conv5"
    top: "conv5"
    name: "relu5"
    type: "ReLU"
}

layer {
    bottom: "conv5"
    top: "conv6"
    name: "conv6"
    type: "Convolution"
    convolution_param {
        num_output: 64
        kernel_size: 3
        pad: 1
        stride: 1
        weight_filler {
            type: "msra"
        }
        
    }
}

layer {
    bottom: "res2a"
    bottom: "conv6"
    top: "res2b"
    name: "res2b"
    type: "Eltwise"
    eltwise_param {
        operation: SUM
    }
}

layer {
    bottom: "res2b"
    top: "res2b"
    name: "res2b_relu"
    type: "ReLU"
}

layer {
    bottom: "res2b"
    top: "conv7"
    name: "conv7"
    type: "Convolution"
    convolution_param {
        num_output: 64
        kernel_size: 1
        pad: 0
        stride: 1
        weight_filler {
            type: "msra"
        }
        

    }
}

layer {
    bottom: "res2b"
    top: "conv8"
    name: "conv8"
    type: "Convolution"
    convolution_param {
        num_output: 64
        kernel_size: 3
        pad: 1
        stride: 1
        weight_filler {
            type: "msra"
        }
        
    }
}

layer {
    bottom: "conv8"
    top: "conv8"
    name: "conv8_relu"
    type: "ReLU"
}

layer {
    bottom: "conv8"
    top: "conv9"
    name: "conv9"
    type: "Convolution"
    convolution_param {
        num_output: 64
        kernel_size: 3
        pad: 1
        stride: 1
        weight_filler {
            type: "msra"
        }
        
    }
}

layer {
    bottom: "conv7"
    bottom: "conv9"
    top: "res3a"
    name: "res3a"
    type: "Eltwise"
    eltwise_param {
        operation: SUM
    }
}

layer {
    bottom: "res3a"
    top: "res3a"
    name: "res3a_relu"
    type: "ReLU"
}

layer {
    bottom: "res3a"
    top: "conv10"
    name: "conv10"
    type: "Convolution"
    convolution_param {
        num_output: 64
        kernel_size: 3
        pad: 1
        stride: 1
        weight_filler {
            type: "msra"
        }
        

    }
}

layer {
    bottom: "conv10"
    top: "conv10"
    name: "relu10"
    type: "ReLU"
}

layer {
    bottom: "conv10"
    top: "conv11"
    name: "conv11"
    type: "Convolution"
    convolution_param {
        num_output: 64
        kernel_size: 3
        pad: 1
        stride: 1
        weight_filler {
            type: "msra"
        }
        
    }
}

layer {
    bottom: "res3a"
    bottom: "conv11"
    top: "res3b"
    name: "res3b"
    type: "Eltwise"
    eltwise_param {
        operation: SUM
    }
}

layer {
    bottom: "res3b"
    top: "res3b"
    name: "res3b_relu"
    type: "ReLU"
}

layer {
    bottom: "res3b"
    top: "conv12"
    name: "conv12"
    type: "Convolution"
    convolution_param {
        num_output: 64
        kernel_size: 1
        pad: 0
        stride: 1
        weight_filler {
            type: "msra"
        }
        

    }
}

layer {
    bottom: "res3b"
    top: "conv13"
    name: "conv13"
    type: "Convolution"
    convolution_param {
        num_output: 64
        kernel_size: 3
        pad: 1
        stride: 1
        weight_filler {
            type: "msra"
        }
        
    }
}

layer {
    bottom: "conv13"
    top: "conv13"
    name: "conv13_relu"
    type: "ReLU"
}

layer {
    bottom: "conv13"
    top: "conv14"
    name: "conv14"
    type: "Convolution"
    convolution_param {
        num_output: 64
        kernel_size: 3
        pad: 1
        stride: 1
        weight_filler {
            type: "msra"
        }
        
    }
}

layer {
    bottom: "conv12"
    bottom: "conv14"
    top: "res4a"
    name: "res4a"
    type: "Eltwise"
    eltwise_param {
        operation: SUM
    }
}

layer {
    bottom: "res4a"
    top: "res4a"
    name: "res4a_relu"
    type: "ReLU"
}

layer {
    bottom: "res4a"
    top: "conv15"
    name: "conv15"
    type: "Convolution"
    convolution_param {
        num_output: 64
        kernel_size: 3
        pad: 1
        stride: 1
        weight_filler {
            type: "msra"
        }
        

    }
}

layer {
    bottom: "conv15"
    top: "conv15"
    name: "relu15"
    type: "ReLU"
}

layer {
    bottom: "conv15"
    top: "conv16"
    name: "conv16"
    type: "Convolution"
    convolution_param {
        num_output: 64
        kernel_size: 3
        pad: 1
        stride: 1
        weight_filler {
            type: "msra"
        }
        
    }
}

layer {
    bottom: "res4a"
    bottom: "conv16"
    top: "res4b"
    name: "res4b"
    type: "Eltwise"
    eltwise_param {
        operation: SUM
    }
}

layer {
    bottom: "res4b"
    top: "res4b"
    name: "res4b_relu"
    type: "ReLU"
}

layer {
  name: "conv17"
  type: "Convolution"
  bottom: "res4b"
  top: "conv17"
  param {
    lr_mult: 1
  }
  param {
    lr_mult: 0.1
  }
  convolution_param {
    num_output: 1
    kernel_size: 3
    stride: 1
    pad: 1
    weight_filler {
      type: "msra"
    }
    
  }
}

layer {
  name: "ssimloss"
  type: "SSIMLoss"
  bottom: "conv17"
  bottom: "label"
  top: "loss"  
  loss_weight: 1             # <- set whatever you fancy
  ssim_loss_param{
    kernel_size: 8           # <- The kernel size is linked to the gaussian variance (circular). The kernel encloses +/1 3*sigma 
    stride: 8                # <- Equal strides in both dimensions
    c1: 0.0001               # <- Let these be
    c2: 0.001                # <- Let these be
  }
}