SemanticFusion编译及运行过程中的一些问题

根据SemanticFusion的要求,搭建好相关的运行环境。(DeConvNet(convolution-deconvolution结构的神经网络) + ElasticFusion(稠密SLAM) + CRF融合  )

使用Kinect2相机运行

./SemanticFusion

输出过程如下:

WARNING: Logging before InitGoogleLogging() is written to STDERR
I0115 21:46:40.668043  7366 net.cpp:46] Initializing net from parameters: 
name: "DeConvNet"
input: "data"
state {
  phase: TEST
}
input_shape {
  dim: 1
  dim: 4
  dim: 224
  dim: 224
}
layer {
  name: "conv1_1"
  type: "Convolution"
  bottom: "data"
  top: "conv1_1"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 64
    pad: 1
    kernel_size: 3
  }
}
layer {
  name: "bn1_1"
  type: "BN"
  bottom: "conv1_1"
  top: "conv1_1"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "relu1_1"
  type: "ReLU"
  bottom: "conv1_1"
  top: "conv1_1"
}
layer {
  name: "conv1_2"
  type: "Convolution"
  bottom: "conv1_1"
  top: "conv1_2"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 64
    pad: 1
    kernel_size: 3
  }
}
layer {
  name: "bn1_2"
  type: "BN"
  bottom: "conv1_2"
  top: "conv1_2"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "relu1_2"
  type: "ReLU"
  bottom: "conv1_2"
  top: "conv1_2"
}
layer {
  name: "pool1"
  type: "Pooling"
  bottom: "conv1_2"
  top: "pool1"
  top: "pool1_mask"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "conv2_1"
  type: "Convolution"
  bottom: "pool1"
  top: "conv2_1"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 128
    pad: 1
    kernel_size: 3
  }
}
layer {
  name: "bn2_1"
  type: "BN"
  bottom: "conv2_1"
  top: "conv2_1"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "relu2_1"
  type: "ReLU"
  bottom: "conv2_1"
  top: "conv2_1"
}
layer {
  name: "conv2_2"
  type: "Convolution"
  bottom: "conv2_1"
  top: "conv2_2"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 128
    pad: 1
    kernel_size: 3
  }
}
layer {
  name: "bn2_2"
  type: "BN"
  bottom: "conv2_2"
  top: "conv2_2"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "relu2_2"
  type: "ReLU"
  bottom: "conv2_2"
  top: "conv2_2"
}
layer {
  name: "pool2"
  type: "Pooling"
  bottom: "conv2_2"
  top: "pool2"
  top: "pool2_mask"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "conv3_1"
  type: "Convolution"
  bottom: "pool2"
  top: "conv3_1"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
  }
}
layer {
  name: "bn3_1"
  type: "BN"
  bottom: "conv3_1"
  top: "conv3_1"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "relu3_1"
  type: "ReLU"
  bottom: "conv3_1"
  top: "conv3_1"
}
layer {
  name: "conv3_2"
  type: "Convolution"
  bottom: "conv3_1"
  top: "conv3_2"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
  }
}
layer {
  name: "bn3_2"
  type: "BN"
  bottom: "conv3_2"
  top: "conv3_2"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "relu3_2"
  type: "ReLU"
  bottom: "conv3_2"
  top: "conv3_2"
}
layer {
  name: "conv3_3"
  type: "Convolution"
  bottom: "conv3_2"
  top: "conv3_3"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
  }
}
layer {
  name: "bn3_3"
  type: "BN"
  bottom: "conv3_3"
  top: "conv3_3"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "relu3_3"
  type: "ReLU"
  bottom: "conv3_3"
  top: "conv3_3"
}
layer {
  name: "pool3"
  type: "Pooling"
  bottom: "conv3_3"
  top: "pool3"
  top: "pool3_mask"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "conv4_1"
  type: "Convolution"
  bottom: "pool3"
  top: "conv4_1"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
  }
}
layer {
  name: "bn4_1"
  type: "BN"
  bottom: "conv4_1"
  top: "conv4_1"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "relu4_1"
  type: "ReLU"
  bottom: "conv4_1"
  top: "conv4_1"
}
layer {
  name: "conv4_2"
  type: "Convolution"
  bottom: "conv4_1"
  top: "conv4_2"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
  }
}
layer {
  name: "bn4_2"
  type: "BN"
  bottom: "conv4_2"
  top: "conv4_2"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "relu4_2"
  type: "ReLU"
  bottom: "conv4_2"
  top: "conv4_2"
}
layer {
  name: "conv4_3"
  type: "Convolution"
  bottom: "conv4_2"
  top: "conv4_3"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
  }
}
layer {
  name: "bn4_3"
  type: "BN"
  bottom: "conv4_3"
  top: "conv4_3"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "relu4_3"
  type: "ReLU"
  bottom: "conv4_3"
  top: "conv4_3"
}
layer {
  name: "pool4"
  type: "Pooling"
  bottom: "conv4_3"
  top: "pool4"
  top: "pool4_mask"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "conv5_1"
  type: "Convolution"
  bottom: "pool4"
  top: "conv5_1"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
  }
}
layer {
  name: "bn5_1"
  type: "BN"
  bottom: "conv5_1"
  top: "conv5_1"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "relu5_1"
  type: "ReLU"
  bottom: "conv5_1"
  top: "conv5_1"
}
layer {
  name: "conv5_2"
  type: "Convolution"
  bottom: "conv5_1"
  top: "conv5_2"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
  }
}
layer {
  name: "bn5_2"
  type: "BN"
  bottom: "conv5_2"
  top: "conv5_2"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "relu5_2"
  type: "ReLU"
  bottom: "conv5_2"
  top: "conv5_2"
}
layer {
  name: "conv5_3"
  type: "Convolution"
  bottom: "conv5_2"
  top: "conv5_3"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
  }
}
layer {
  name: "bn5_3"
  type: "BN"
  bottom: "conv5_3"
  top: "conv5_3"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "relu5_3"
  type: "ReLU"
  bottom: "conv5_3"
  top: "conv5_3"
}
layer {
  name: "pool5"
  type: "Pooling"
  bottom: "conv5_3"
  top: "pool5"
  top: "pool5_mask"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "fc6"
  type: "Convolution"
  bottom: "pool5"
  top: "fc6"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 4096
    kernel_size: 7
  }
}
layer {
  name: "bnfc6"
  type: "BN"
  bottom: "fc6"
  top: "fc6"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "relu6"
  type: "ReLU"
  bottom: "fc6"
  top: "fc6"
}
layer {
  name: "fc7"
  type: "Convolution"
  bottom: "fc6"
  top: "fc7"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 4096
    kernel_size: 1
  }
}
layer {
  name: "bnfc7"
  type: "BN"
  bottom: "fc7"
  top: "fc7"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "relu7"
  type: "ReLU"
  bottom: "fc7"
  top: "fc7"
}
layer {
  name: "fc6-deconv"
  type: "Deconvolution"
  bottom: "fc7"
  top: "fc6-deconv"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 512
    kernel_size: 7
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "fc6-deconv-bn"
  type: "BN"
  bottom: "fc6-deconv"
  top: "fc6-deconv"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "fc6-deconv-relu"
  type: "ReLU"
  bottom: "fc6-deconv"
  top: "fc6-deconv"
}
layer {
  name: "unpool5"
  type: "UnPooling"
  bottom: "fc6-deconv"
  bottom: "pool5_mask"
  top: "unpool5"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "deconv5_1"
  type: "Deconvolution"
  bottom: "unpool5"
  top: "deconv5_1"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "debn5_1"
  type: "BN"
  bottom: "deconv5_1"
  top: "deconv5_1"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "derelu5_1"
  type: "ReLU"
  bottom: "deconv5_1"
  top: "deconv5_1"
}
layer {
  name: "deconv5_2"
  type: "Deconvolution"
  bottom: "deconv5_1"
  top: "deconv5_2"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "debn5_2"
  type: "BN"
  bottom: "deconv5_2"
  top: "deconv5_2"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "derelu5_2"
  type: "ReLU"
  bottom: "deconv5_2"
  top: "deconv5_2"
}
layer {
  name: "deconv5_3"
  type: "Deconvolution"
  bottom: "deconv5_2"
  top: "deconv5_3"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "debn5_3"
  type: "BN"
  bottom: "deconv5_3"
  top: "deconv5_3"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "derelu5_3"
  type: "ReLU"
  bottom: "deconv5_3"
  top: "deconv5_3"
}
layer {
  name: "unpool4"
  type: "UnPooling"
  bottom: "deconv5_3"
  bottom: "pool4_mask"
  top: "unpool4"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "deconv4_1"
  type: "Deconvolution"
  bottom: "unpool4"
  top: "deconv4_1"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "debn4_1"
  type: "BN"
  bottom: "deconv4_1"
  top: "deconv4_1"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "derelu4_1"
  type: "ReLU"
  bottom: "deconv4_1"
  top: "deconv4_1"
}
layer {
  name: "deconv4_2"
  type: "Deconvolution"
  bottom: "deconv4_1"
  top: "deconv4_2"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "debn4_2"
  type: "BN"
  bottom: "deconv4_2"
  top: "deconv4_2"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "derelu4_2"
  type: "ReLU"
  bottom: "deconv4_2"
  top: "deconv4_2"
}
layer {
  name: "deconv4_3"
  type: "Deconvolution"
  bottom: "deconv4_2"
  top: "deconv4_3"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "debn4_3"
  type: "BN"
  bottom: "deconv4_3"
  top: "deconv4_3"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "derelu4_3"
  type: "ReLU"
  bottom: "deconv4_3"
  top: "deconv4_3"
}
layer {
  name: "unpool3"
  type: "UnPooling"
  bottom: "deconv4_3"
  bottom: "pool3_mask"
  top: "unpool3"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "deconv3_1"
  type: "Deconvolution"
  bottom: "unpool3"
  top: "deconv3_1"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "debn3_1"
  type: "BN"
  bottom: "deconv3_1"
  top: "deconv3_1"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "derelu3_1"
  type: "ReLU"
  bottom: "deconv3_1"
  top: "deconv3_1"
}
layer {
  name: "deconv3_2"
  type: "Deconvolution"
  bottom: "deconv3_1"
  top: "deconv3_2"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "debn3_2"
  type: "BN"
  bottom: "deconv3_2"
  top: "deconv3_2"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "derelu3_2"
  type: "ReLU"
  bottom: "deconv3_2"
  top: "deconv3_2"
}
layer {
  name: "deconv3_3"
  type: "Deconvolution"
  bottom: "deconv3_2"
  top: "deconv3_3"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 128
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "debn3_3"
  type: "BN"
  bottom: "deconv3_3"
  top: "deconv3_3"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "derelu3_3"
  type: "ReLU"
  bottom: "deconv3_3"
  top: "deconv3_3"
}
layer {
  name: "unpool2"
  type: "UnPooling"
  bottom: "deconv3_3"
  bottom: "pool2_mask"
  top: "unpool2"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "deconv2_1"
  type: "Deconvolution"
  bottom: "unpool2"
  top: "deconv2_1"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 128
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "debn2_1"
  type: "BN"
  bottom: "deconv2_1"
  top: "deconv2_1"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "derelu2_1"
  type: "ReLU"
  bottom: "deconv2_1"
  top: "deconv2_1"
}
layer {
  name: "deconv2_2"
  type: "Deconvolution"
  bottom: "deconv2_1"
  top: "deconv2_2"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 64
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "debn2_2"
  type: "BN"
  bottom: "deconv2_2"
  top: "deconv2_2"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "derelu2_2"
  type: "ReLU"
  bottom: "deconv2_2"
  top: "deconv2_2"
}
layer {
  name: "unpool1"
  type: "UnPooling"
  bottom: "deconv2_2"
  bottom: "pool1_mask"
  top: "unpool1"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "deconv1_1"
  type: "Deconvolution"
  bottom: "unpool1"
  top: "deconv1_1"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 64
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "debn1_1"
  type: "BN"
  bottom: "deconv1_1"
  top: "deconv1_1"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "derelu1_1"
  type: "ReLU"
  bottom: "deconv1_1"
  top: "deconv1_1"
}
layer {
  name: "deconv1_2"
  type: "Deconvolution"
  bottom: "deconv1_1"
  top: "deconv1_2"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 64
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "debn1_2"
  type: "BN"
  bottom: "deconv1_2"
  top: "deconv1_2"
  bn_param {
    scale_filler {
      type: "constant"
      value: 1
    }
    shift_filler {
      type: "constant"
      value: 0.001
    }
    bn_mode: INFERENCE
  }
}
layer {
  name: "derelu1_2"
  type: "ReLU"
  bottom: "deconv1_2"
  top: "deconv1_2"
}
layer {
  name: "class_score_nyu"
  type: "Convolution"
  bottom: "deconv1_2"
  top: "class_score"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 14
    kernel_size: 1
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "probability"
  type: "Softmax"
  bottom: "class_score"
  top: "probability"
  softmax_param {
    axis: 1
  }
}
I0115 21:46:40.668938  7366 net.cpp:410] Input 0 -> data
I0115 21:46:40.674125  7366 layer_factory.hpp:77] Creating layer conv1_1
I0115 21:46:40.674154  7366 net.cpp:103] Creating Layer conv1_1
I0115 21:46:40.674160  7366 net.cpp:451] conv1_1 <- data
I0115 21:46:40.674166  7366 net.cpp:408] conv1_1 -> conv1_1
I0115 21:46:40.675429  7366 net.cpp:147] Setting up conv1_1
I0115 21:46:40.675458  7366 net.cpp:154] Top shape: 1 64 224 224 (3211264)
I0115 21:46:40.675474  7366 net.cpp:162] Memory required for data: 12845056
I0115 21:46:40.675510  7366 layer_factory.hpp:77] Creating layer bn1_1
I0115 21:46:40.675518  7366 net.cpp:103] Creating Layer bn1_1
I0115 21:46:40.675523  7366 net.cpp:451] bn1_1 <- conv1_1
I0115 21:46:40.675525  7366 net.cpp:394] bn1_1 -> conv1_1 (in-place)
I0115 21:46:40.675704  7366 net.cpp:147] Setting up bn1_1
I0115 21:46:40.675710  7366 net.cpp:154] Top shape: 1 64 224 224 (3211264)
I0115 21:46:40.675731  7366 net.cpp:162] Memory required for data: 25690112
I0115 21:46:40.675737  7366 layer_factory.hpp:77] Creating layer relu1_1
I0115 21:46:40.675741  7366 net.cpp:103] Creating Layer relu1_1
I0115 21:46:40.675745  7366 net.cpp:451] relu1_1 <- conv1_1
I0115 21:46:40.675750  7366 net.cpp:394] relu1_1 -> conv1_1 (in-place)
I0115 21:46:40.675755  7366 net.cpp:147] Setting up relu1_1
I0115 21:46:40.675758  7366 net.cpp:154] Top shape: 1 64 224 224 (3211264)
I0115 21:46:40.675761  7366 net.cpp:162] Memory required for data: 38535168
I0115 21:46:40.675765  7366 layer_factory.hpp:77] Creating layer conv1_2
I0115 21:46:40.675770  7366 net.cpp:103] Creating Layer conv1_2
I0115 21:46:40.675772  7366 net.cpp:451] conv1_2 <- conv1_1
I0115 21:46:40.675776  7366 net.cpp:408] conv1_2 -> conv1_2
I0115 21:46:40.675914  7366 net.cpp:147] Setting up conv1_2
I0115 21:46:40.675920  7366 net.cpp:154] Top shape: 1 64 224 224 (3211264)
I0115 21:46:40.675922  7366 net.cpp:162] Memory required for data: 51380224
I0115 21:46:40.675927  7366 layer_factory.hpp:77] Creating layer bn1_2
I0115 21:46:40.675933  7366 net.cpp:103] Creating Layer bn1_2
I0115 21:46:40.675937  7366 net.cpp:451] bn1_2 <- conv1_2
I0115 21:46:40.675941  7366 net.cpp:394] bn1_2 -> conv1_2 (in-place)
I0115 21:46:40.676064  7366 net.cpp:147] Setting up bn1_2
I0115 21:46:40.676069  7366 net.cpp:154] Top shape: 1 64 224 224 (3211264)
I0115 21:46:40.676072  7366 net.cpp:162] Memory required for data: 64225280
I0115 21:46:40.676077  7366 layer_factory.hpp:77] Creating layer relu1_2
I0115 21:46:40.676084  7366 net.cpp:103] Creating Layer relu1_2
I0115 21:46:40.676087  7366 net.cpp:451] relu1_2 <- conv1_2
I0115 21:46:40.676090  7366 net.cpp:394] relu1_2 -> conv1_2 (in-place)
I0115 21:46:40.676095  7366 net.cpp:147] Setting up relu1_2
I0115 21:46:40.676098  7366 net.cpp:154] Top shape: 1 64 224 224 (3211264)
I0115 21:46:40.676100  7366 net.cpp:162] Memory required for data: 77070336
I0115 21:46:40.676103  7366 layer_factory.hpp:77] Creating layer pool1
I0115 21:46:40.676107  7366 net.cpp:103] Creating Layer pool1
I0115 21:46:40.676110  7366 net.cpp:451] pool1 <- conv1_2
I0115 21:46:40.676115  7366 net.cpp:408] pool1 -> pool1
I0115 21:46:40.676120  7366 net.cpp:408] pool1 -> pool1_mask
I0115 21:46:40.676147  7366 net.cpp:147] Setting up pool1
I0115 21:46:40.676152  7366 net.cpp:154] Top shape: 1 64 112 112 (802816)
I0115 21:46:40.676156  7366 net.cpp:154] Top shape: 1 64 112 112 (802816)
I0115 21:46:40.676159  7366 net.cpp:162] Memory required for data: 83492864
I0115 21:46:40.676162  7366 layer_factory.hpp:77] Creating layer conv2_1
I0115 21:46:40.676167  7366 net.cpp:103] Creating Layer conv2_1
I0115 21:46:40.676170  7366 net.cpp:451] conv2_1 <- pool1
I0115 21:46:40.676179  7366 net.cpp:408] conv2_1 -> conv2_1
I0115 21:46:40.677409  7366 net.cpp:147] Setting up conv2_1
I0115 21:46:40.677420  7366 net.cpp:154] Top shape: 1 128 112 112 (1605632)
I0115 21:46:40.677423  7366 net.cpp:162] Memory required for data: 89915392
I0115 21:46:40.677433  7366 layer_factory.hpp:77] Creating layer bn2_1
I0115 21:46:40.677439  7366 net.cpp:103] Creating Layer bn2_1
I0115 21:46:40.677443  7366 net.cpp:451] bn2_1 <- conv2_1
I0115 21:46:40.677448  7366 net.cpp:394] bn2_1 -> conv2_1 (in-place)
I0115 21:46:40.677558  7366 net.cpp:147] Setting up bn2_1
I0115 21:46:40.677564  7366 net.cpp:154] Top shape: 1 128 112 112 (1605632)
I0115 21:46:40.677567  7366 net.cpp:162] Memory required for data: 96337920
I0115 21:46:40.678447  7366 layer_factory.hpp:77] Creating layer relu2_1
I0115 21:46:40.678452  7366 net.cpp:103] Creating Layer relu2_1
I0115 21:46:40.678454  7366 net.cpp:451] relu2_1 <- conv2_1
I0115 21:46:40.678458  7366 net.cpp:394] relu2_1 -> conv2_1 (in-place)
I0115 21:46:40.678462  7366 net.cpp:147] Setting up relu2_1
I0115 21:46:40.678467  7366 net.cpp:154] Top shape: 1 128 112 112 (1605632)
I0115 21:46:40.678469  7366 net.cpp:162] Memory required for data: 102760448
I0115 21:46:40.678472  7366 layer_factory.hpp:77] Creating layer conv2_2
I0115 21:46:40.678478  7366 net.cpp:103] Creating Layer conv2_2
I0115 21:46:40.678479  7366 net.cpp:451] conv2_2 <- conv2_1
I0115 21:46:40.678484  7366 net.cpp:408] conv2_2 -> conv2_2
I0115 21:46:40.678643  7366 net.cpp:147] Setting up conv2_2
I0115 21:46:40.678649  7366 net.cpp:154] Top shape: 1 128 112 112 (1605632)
I0115 21:46:40.678653  7366 net.cpp:162] Memory required for data: 109182976
I0115 21:46:40.678656  7366 layer_factory.hpp:77] Creating layer bn2_2
I0115 21:46:40.678661  7366 net.cpp:103] Creating Layer bn2_2
I0115 21:46:40.678664  7366 net.cpp:451] bn2_2 <- conv2_2
I0115 21:46:40.678668  7366 net.cpp:394] bn2_2 -> conv2_2 (in-place)
I0115 21:46:40.678774  7366 net.cpp:147] Setting up bn2_2
I0115 21:46:40.678779  7366 net.cpp:154] Top shape: 1 128 112 112 (1605632)
I0115 21:46:40.678782  7366 net.cpp:162] Memory required for data: 115605504
I0115 21:46:40.678786  7366 layer_factory.hpp:77] Creating layer relu2_2
I0115 21:46:40.678791  7366 net.cpp:103] Creating Layer relu2_2
I0115 21:46:40.678793  7366 net.cpp:451] relu2_2 <- conv2_2
I0115 21:46:40.678797  7366 net.cpp:394] relu2_2 -> conv2_2 (in-place)
I0115 21:46:40.678800  7366 net.cpp:147] Setting up relu2_2
I0115 21:46:40.678804  7366 net.cpp:154] Top shape: 1 128 112 112 (1605632)
I0115 21:46:40.678807  7366 net.cpp:162] Memory required for data: 122028032
I0115 21:46:40.678809  7366 layer_factory.hpp:77] Creating layer pool2
I0115 21:46:40.678814  7366 net.cpp:103] Creating Layer pool2
I0115 21:46:40.678817  7366 net.cpp:451] pool2 <- conv2_2
I0115 21:46:40.678822  7366 net.cpp:408] pool2 -> pool2
I0115 21:46:40.678827  7366 net.cpp:408] pool2 -> pool2_mask
I0115 21:46:40.678848  7366 net.cpp:147] Setting up pool2
I0115 21:46:40.678853  7366 net.cpp:154] Top shape: 1 128 56 56 (401408)
I0115 21:46:40.678858  7366 net.cpp:154] Top shape: 1 128 56 56 (401408)
I0115 21:46:40.678859  7366 net.cpp:162] Memory required for data: 125239296
I0115 21:46:40.678863  7366 layer_factory.hpp:77] Creating layer conv3_1
I0115 21:46:40.678869  7366 net.cpp:103] Creating Layer conv3_1
I0115 21:46:40.678871  7366 net.cpp:451] conv3_1 <- pool2
I0115 21:46:40.678875  7366 net.cpp:408] conv3_1 -> conv3_1
I0115 21:46:40.679061  7366 net.cpp:147] Setting up conv3_1
I0115 21:46:40.679067  7366 net.cpp:154] Top shape: 1 256 56 56 (802816)
I0115 21:46:40.679070  7366 net.cpp:162] Memory required for data: 128450560
I0115 21:46:40.679077  7366 layer_factory.hpp:77] Creating layer bn3_1
I0115 21:46:40.679081  7366 net.cpp:103] Creating Layer bn3_1
I0115 21:46:40.679085  7366 net.cpp:451] bn3_1 <- conv3_1
I0115 21:46:40.679090  7366 net.cpp:394] bn3_1 -> conv3_1 (in-place)
I0115 21:46:40.679199  7366 net.cpp:147] Setting up bn3_1
I0115 21:46:40.679205  7366 net.cpp:154] Top shape: 1 256 56 56 (802816)
I0115 21:46:40.679208  7366 net.cpp:162] Memory required for data: 131661824
I0115 21:46:40.679213  7366 layer_factory.hpp:77] Creating layer relu3_1
I0115 21:46:40.679219  7366 net.cpp:103] Creating Layer relu3_1
I0115 21:46:40.679221  7366 net.cpp:451] relu3_1 <- conv3_1
I0115 21:46:40.679224  7366 net.cpp:394] relu3_1 -> conv3_1 (in-place)
I0115 21:46:40.679229  7366 net.cpp:147] Setting up relu3_1
I0115 21:46:40.679232  7366 net.cpp:154] Top shape: 1 256 56 56 (802816)
I0115 21:46:40.679235  7366 net.cpp:162] Memory required for data: 134873088
I0115 21:46:40.679239  7366 layer_factory.hpp:77] Creating layer conv3_2
I0115 21:46:40.679242  7366 net.cpp:103] Creating Layer conv3_2
I0115 21:46:40.679245  7366 net.cpp:451] conv3_2 <- conv3_1
I0115 21:46:40.679250  7366 net.cpp:408] conv3_2 -> conv3_2
I0115 21:46:40.680331  7366 net.cpp:147] Setting up conv3_2
I0115 21:46:40.680361  7366 net.cpp:154] Top shape: 1 256 56 56 (802816)
I0115 21:46:40.680392  7366 net.cpp:162] Memory required for data: 138084352
I0115 21:46:40.680418  7366 layer_factory.hpp:77] Creating layer bn3_2
I0115 21:46:40.680450  7366 net.cpp:103] Creating Layer bn3_2
I0115 21:46:40.680456  7366 net.cpp:451] bn3_2 <- conv3_2
I0115 21:46:40.680466  7366 net.cpp:394] bn3_2 -> conv3_2 (in-place)
I0115 21:46:40.680661  7366 net.cpp:147] Setting up bn3_2
I0115 21:46:40.680670  7366 net.cpp:154] Top shape: 1 256 56 56 (802816)
I0115 21:46:40.680672  7366 net.cpp:162] Memory required for data: 141295616
I0115 21:46:40.680676  7366 layer_factory.hpp:77] Creating layer relu3_2
I0115 21:46:40.680680  7366 net.cpp:103] Creating Layer relu3_2
I0115 21:46:40.680683  7366 net.cpp:451] relu3_2 <- conv3_2
I0115 21:46:40.680686  7366 net.cpp:394] relu3_2 -> conv3_2 (in-place)
I0115 21:46:40.680691  7366 net.cpp:147] Setting up relu3_2
I0115 21:46:40.680693  7366 net.cpp:154] Top shape: 1 256 56 56 (802816)
I0115 21:46:40.680696  7366 net.cpp:162] Memory required for data: 144506880
I0115 21:46:40.680698  7366 layer_factory.hpp:77] Creating layer conv3_3
I0115 21:46:40.680702  7366 net.cpp:103] Creating Layer conv3_3
I0115 21:46:40.680706  7366 net.cpp:451] conv3_3 <- conv3_2
I0115 21:46:40.680711  7366 net.cpp:408] conv3_3 -> conv3_3
I0115 21:46:40.681591  7366 net.cpp:147] Setting up conv3_3
I0115 21:46:40.681601  7366 net.cpp:154] Top shape: 1 256 56 56 (802816)
I0115 21:46:40.681603  7366 net.cpp:162] Memory required for data: 147718144
I0115 21:46:40.681607  7366 layer_factory.hpp:77] Creating layer bn3_3
I0115 21:46:40.681612  7366 net.cpp:103] Creating Layer bn3_3
I0115 21:46:40.681615  7366 net.cpp:451] bn3_3 <- conv3_3
I0115 21:46:40.681618  7366 net.cpp:394] bn3_3 -> conv3_3 (in-place)
I0115 21:46:40.681782  7366 net.cpp:147] Setting up bn3_3
I0115 21:46:40.681787  7366 net.cpp:154] Top shape: 1 256 56 56 (802816)
I0115 21:46:40.681790  7366 net.cpp:162] Memory required for data: 150929408
I0115 21:46:40.681793  7366 layer_factory.hpp:77] Creating layer relu3_3
I0115 21:46:40.681797  7366 net.cpp:103] Creating Layer relu3_3
I0115 21:46:40.681799  7366 net.cpp:451] relu3_3 <- conv3_3
I0115 21:46:40.681802  7366 net.cpp:394] relu3_3 -> conv3_3 (in-place)
I0115 21:46:40.681807  7366 net.cpp:147] Setting up relu3_3
I0115 21:46:40.681809  7366 net.cpp:154] Top shape: 1 256 56 56 (802816)
I0115 21:46:40.681812  7366 net.cpp:162] Memory required for data: 154140672
I0115 21:46:40.681813  7366 layer_factory.hpp:77] Creating layer pool3
I0115 21:46:40.681819  7366 net.cpp:103] Creating Layer pool3
I0115 21:46:40.681823  7366 net.cpp:451] pool3 <- conv3_3
I0115 21:46:40.681826  7366 net.cpp:408] pool3 -> pool3
I0115 21:46:40.681831  7366 net.cpp:408] pool3 -> pool3_mask
I0115 21:46:40.681903  7366 net.cpp:147] Setting up pool3
I0115 21:46:40.681907  7366 net.cpp:154] Top shape: 1 256 28 28 (200704)
I0115 21:46:40.681911  7366 net.cpp:154] Top shape: 1 256 28 28 (200704)
I0115 21:46:40.681933  7366 net.cpp:162] Memory required for data: 155746304
I0115 21:46:40.681936  7366 layer_factory.hpp:77] Creating layer conv4_1
I0115 21:46:40.681942  7366 net.cpp:103] Creating Layer conv4_1
I0115 21:46:40.681946  7366 net.cpp:451] conv4_1 <- pool3
I0115 21:46:40.681951  7366 net.cpp:408] conv4_1 -> conv4_1
I0115 21:46:40.684136  7366 net.cpp:147] Setting up conv4_1
I0115 21:46:40.684155  7366 net.cpp:154] Top shape: 1 512 28 28 (401408)
I0115 21:46:40.684157  7366 net.cpp:162] Memory required for data: 157351936
I0115 21:46:40.684164  7366 layer_factory.hpp:77] Creating layer bn4_1
I0115 21:46:40.684190  7366 net.cpp:103] Creating Layer bn4_1
I0115 21:46:40.684195  7366 net.cpp:451] bn4_1 <- conv4_1
I0115 21:46:40.684201  7366 net.cpp:394] bn4_1 -> conv4_1 (in-place)
I0115 21:46:40.684345  7366 net.cpp:147] Setting up bn4_1
I0115 21:46:40.684350  7366 net.cpp:154] Top shape: 1 512 28 28 (401408)
I0115 21:46:40.684371  7366 net.cpp:162] Memory required for data: 158957568
I0115 21:46:40.684376  7366 layer_factory.hpp:77] Creating layer relu4_1
I0115 21:46:40.684381  7366 net.cpp:103] Creating Layer relu4_1
I0115 21:46:40.684384  7366 net.cpp:451] relu4_1 <- conv4_1
I0115 21:46:40.684388  7366 net.cpp:394] relu4_1 -> conv4_1 (in-place)
I0115 21:46:40.684392  7366 net.cpp:147] Setting up relu4_1
I0115 21:46:40.684396  7366 net.cpp:154] Top shape: 1 512 28 28 (401408)
I0115 21:46:40.684398  7366 net.cpp:162] Memory required for data: 160563200
I0115 21:46:40.684401  7366 layer_factory.hpp:77] Creating layer conv4_2
I0115 21:46:40.684406  7366 net.cpp:103] Creating Layer conv4_2
I0115 21:46:40.684409  7366 net.cpp:451] conv4_2 <- conv4_1
I0115 21:46:40.684414  7366 net.cpp:408] conv4_2 -> conv4_2
I0115 21:46:40.687465  7366 net.cpp:147] Setting up conv4_2
I0115 21:46:40.687495  7366 net.cpp:154] Top shape: 1 512 28 28 (401408)
I0115 21:46:40.687499  7366 net.cpp:162] Memory required for data: 162168832
I0115 21:46:40.687511  7366 layer_factory.hpp:77] Creating layer bn4_2
I0115 21:46:40.687521  7366 net.cpp:103] Creating Layer bn4_2
I0115 21:46:40.687526  7366 net.cpp:451] bn4_2 <- conv4_2
I0115 21:46:40.687536  7366 net.cpp:394] bn4_2 -> conv4_2 (in-place)
I0115 21:46:40.687655  7366 net.cpp:147] Setting up bn4_2
I0115 21:46:40.687660  7366 net.cpp:154] Top shape: 1 512 28 28 (401408)
I0115 21:46:40.687664  7366 net.cpp:162] Memory required for data: 163774464
I0115 21:46:40.687667  7366 layer_factory.hpp:77] Creating layer relu4_2
I0115 21:46:40.687670  7366 net.cpp:103] Creating Layer relu4_2
I0115 21:46:40.687674  7366 net.cpp:451] relu4_2 <- conv4_2
I0115 21:46:40.687677  7366 net.cpp:394] relu4_2 -> conv4_2 (in-place)
I0115 21:46:40.687681  7366 net.cpp:147] Setting up relu4_2
I0115 21:46:40.687685  7366 net.cpp:154] Top shape: 1 512 28 28 (401408)
I0115 21:46:40.687686  7366 net.cpp:162] Memory required for data: 165380096
I0115 21:46:40.687688  7366 layer_factory.hpp:77] Creating layer conv4_3
I0115 21:46:40.687693  7366 net.cpp:103] Creating Layer conv4_3
I0115 21:46:40.687697  7366 net.cpp:451] conv4_3 <- conv4_2
I0115 21:46:40.687701  7366 net.cpp:408] conv4_3 -> conv4_3
I0115 21:46:40.690676  7366 net.cpp:147] Setting up conv4_3
I0115 21:46:40.690702  7366 net.cpp:154] Top shape: 1 512 28 28 (401408)
I0115 21:46:40.690704  7366 net.cpp:162] Memory required for data: 166985728
I0115 21:46:40.690711  7366 layer_factory.hpp:77] Creating layer bn4_3
I0115 21:46:40.690718  7366 net.cpp:103] Creating Layer bn4_3
I0115 21:46:40.690721  7366 net.cpp:451] bn4_3 <- conv4_3
I0115 21:46:40.690726  7366 net.cpp:394] bn4_3 -> conv4_3 (in-place)
I0115 21:46:40.690865  7366 net.cpp:147] Setting up bn4_3
I0115 21:46:40.690871  7366 net.cpp:154] Top shape: 1 512 28 28 (401408)
I0115 21:46:40.690873  7366 net.cpp:162] Memory required for data: 168591360
I0115 21:46:40.690877  7366 layer_factory.hpp:77] Creating layer relu4_3
I0115 21:46:40.690901  7366 net.cpp:103] Creating Layer relu4_3
I0115 21:46:40.690906  7366 net.cpp:451] relu4_3 <- conv4_3
I0115 21:46:40.690909  7366 net.cpp:394] relu4_3 -> conv4_3 (in-place)
I0115 21:46:40.690914  7366 net.cpp:147] Setting up relu4_3
I0115 21:46:40.690918  7366 net.cpp:154] Top shape: 1 512 28 28 (401408)
I0115 21:46:40.690922  7366 net.cpp:162] Memory required for data: 170196992
I0115 21:46:40.690923  7366 layer_factory.hpp:77] Creating layer pool4
I0115 21:46:40.690929  7366 net.cpp:103] Creating Layer pool4
I0115 21:46:40.690932  7366 net.cpp:451] pool4 <- conv4_3
I0115 21:46:40.690937  7366 net.cpp:408] pool4 -> pool4
I0115 21:46:40.690942  7366 net.cpp:408] pool4 -> pool4_mask
I0115 21:46:40.690964  7366 net.cpp:147] Setting up pool4
I0115 21:46:40.690970  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:40.690973  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:40.690976  7366 net.cpp:162] Memory required for data: 170999808
I0115 21:46:40.690979  7366 layer_factory.hpp:77] Creating layer conv5_1
I0115 21:46:40.690986  7366 net.cpp:103] Creating Layer conv5_1
I0115 21:46:40.690989  7366 net.cpp:451] conv5_1 <- pool4
I0115 21:46:40.690994  7366 net.cpp:408] conv5_1 -> conv5_1
I0115 21:46:40.693842  7366 net.cpp:147] Setting up conv5_1
I0115 21:46:40.693869  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:40.693872  7366 net.cpp:162] Memory required for data: 171401216
I0115 21:46:40.693897  7366 layer_factory.hpp:77] Creating layer bn5_1
I0115 21:46:40.693905  7366 net.cpp:103] Creating Layer bn5_1
I0115 21:46:40.693912  7366 net.cpp:451] bn5_1 <- conv5_1
I0115 21:46:40.693919  7366 net.cpp:394] bn5_1 -> conv5_1 (in-place)
I0115 21:46:40.694041  7366 net.cpp:147] Setting up bn5_1
I0115 21:46:40.694046  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:40.694049  7366 net.cpp:162] Memory required for data: 171802624
I0115 21:46:40.694053  7366 layer_factory.hpp:77] Creating layer relu5_1
I0115 21:46:40.694058  7366 net.cpp:103] Creating Layer relu5_1
I0115 21:46:40.694061  7366 net.cpp:451] relu5_1 <- conv5_1
I0115 21:46:40.694066  7366 net.cpp:394] relu5_1 -> conv5_1 (in-place)
I0115 21:46:40.694070  7366 net.cpp:147] Setting up relu5_1
I0115 21:46:40.694074  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:40.694077  7366 net.cpp:162] Memory required for data: 172204032
I0115 21:46:40.694079  7366 layer_factory.hpp:77] Creating layer conv5_2
I0115 21:46:40.694087  7366 net.cpp:103] Creating Layer conv5_2
I0115 21:46:40.694089  7366 net.cpp:451] conv5_2 <- conv5_1
I0115 21:46:40.694093  7366 net.cpp:408] conv5_2 -> conv5_2
I0115 21:46:40.697716  7366 net.cpp:147] Setting up conv5_2
I0115 21:46:40.697748  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:40.697752  7366 net.cpp:162] Memory required for data: 172605440
I0115 21:46:40.697759  7366 layer_factory.hpp:77] Creating layer bn5_2
I0115 21:46:40.697767  7366 net.cpp:103] Creating Layer bn5_2
I0115 21:46:40.697772  7366 net.cpp:451] bn5_2 <- conv5_2
I0115 21:46:40.697778  7366 net.cpp:394] bn5_2 -> conv5_2 (in-place)
I0115 21:46:40.697917  7366 net.cpp:147] Setting up bn5_2
I0115 21:46:40.697923  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:40.697926  7366 net.cpp:162] Memory required for data: 173006848
I0115 21:46:40.697930  7366 layer_factory.hpp:77] Creating layer relu5_2
I0115 21:46:40.697935  7366 net.cpp:103] Creating Layer relu5_2
I0115 21:46:40.697957  7366 net.cpp:451] relu5_2 <- conv5_2
I0115 21:46:40.697962  7366 net.cpp:394] relu5_2 -> conv5_2 (in-place)
I0115 21:46:40.697965  7366 net.cpp:147] Setting up relu5_2
I0115 21:46:40.697969  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:40.697973  7366 net.cpp:162] Memory required for data: 173408256
I0115 21:46:40.697975  7366 layer_factory.hpp:77] Creating layer conv5_3
I0115 21:46:40.697980  7366 net.cpp:103] Creating Layer conv5_3
I0115 21:46:40.697983  7366 net.cpp:451] conv5_3 <- conv5_2
I0115 21:46:40.697989  7366 net.cpp:408] conv5_3 -> conv5_3
I0115 21:46:40.701063  7366 net.cpp:147] Setting up conv5_3
I0115 21:46:40.701089  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:40.701093  7366 net.cpp:162] Memory required for data: 173809664
I0115 21:46:40.701100  7366 layer_factory.hpp:77] Creating layer bn5_3
I0115 21:46:40.701109  7366 net.cpp:103] Creating Layer bn5_3
I0115 21:46:40.701113  7366 net.cpp:451] bn5_3 <- conv5_3
I0115 21:46:40.701118  7366 net.cpp:394] bn5_3 -> conv5_3 (in-place)
I0115 21:46:40.701244  7366 net.cpp:147] Setting up bn5_3
I0115 21:46:40.701249  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:40.701252  7366 net.cpp:162] Memory required for data: 174211072
I0115 21:46:40.701256  7366 layer_factory.hpp:77] Creating layer relu5_3
I0115 21:46:40.701261  7366 net.cpp:103] Creating Layer relu5_3
I0115 21:46:40.701264  7366 net.cpp:451] relu5_3 <- conv5_3
I0115 21:46:40.701268  7366 net.cpp:394] relu5_3 -> conv5_3 (in-place)
I0115 21:46:40.701272  7366 net.cpp:147] Setting up relu5_3
I0115 21:46:40.701277  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:40.701279  7366 net.cpp:162] Memory required for data: 174612480
I0115 21:46:40.701282  7366 layer_factory.hpp:77] Creating layer pool5
I0115 21:46:40.701287  7366 net.cpp:103] Creating Layer pool5
I0115 21:46:40.701289  7366 net.cpp:451] pool5 <- conv5_3
I0115 21:46:40.701314  7366 net.cpp:408] pool5 -> pool5
I0115 21:46:40.701319  7366 net.cpp:408] pool5 -> pool5_mask
I0115 21:46:40.701359  7366 net.cpp:147] Setting up pool5
I0115 21:46:40.701364  7366 net.cpp:154] Top shape: 1 512 7 7 (25088)
I0115 21:46:40.701367  7366 net.cpp:154] Top shape: 1 512 7 7 (25088)
I0115 21:46:40.701370  7366 net.cpp:162] Memory required for data: 174813184
I0115 21:46:40.701372  7366 layer_factory.hpp:77] Creating layer fc6
I0115 21:46:40.701380  7366 net.cpp:103] Creating Layer fc6
I0115 21:46:40.701381  7366 net.cpp:451] fc6 <- pool5
I0115 21:46:40.701385  7366 net.cpp:408] fc6 -> fc6
I0115 21:46:40.848129  7366 net.cpp:147] Setting up fc6
I0115 21:46:40.848153  7366 net.cpp:154] Top shape: 1 4096 1 1 (4096)
I0115 21:46:40.848156  7366 net.cpp:162] Memory required for data: 174829568
I0115 21:46:40.848162  7366 layer_factory.hpp:77] Creating layer bnfc6
I0115 21:46:40.848170  7366 net.cpp:103] Creating Layer bnfc6
I0115 21:46:40.848175  7366 net.cpp:451] bnfc6 <- fc6
I0115 21:46:40.848179  7366 net.cpp:394] bnfc6 -> fc6 (in-place)
I0115 21:46:40.848307  7366 net.cpp:147] Setting up bnfc6
I0115 21:46:40.848314  7366 net.cpp:154] Top shape: 1 4096 1 1 (4096)
I0115 21:46:40.848315  7366 net.cpp:162] Memory required for data: 174845952
I0115 21:46:40.848320  7366 layer_factory.hpp:77] Creating layer relu6
I0115 21:46:40.848323  7366 net.cpp:103] Creating Layer relu6
I0115 21:46:40.848326  7366 net.cpp:451] relu6 <- fc6
I0115 21:46:40.848330  7366 net.cpp:394] relu6 -> fc6 (in-place)
I0115 21:46:40.848333  7366 net.cpp:147] Setting up relu6
I0115 21:46:40.848336  7366 net.cpp:154] Top shape: 1 4096 1 1 (4096)
I0115 21:46:40.848338  7366 net.cpp:162] Memory required for data: 174862336
I0115 21:46:40.848340  7366 layer_factory.hpp:77] Creating layer fc7
I0115 21:46:40.848346  7366 net.cpp:103] Creating Layer fc7
I0115 21:46:40.848351  7366 net.cpp:451] fc7 <- fc6
I0115 21:46:40.848354  7366 net.cpp:408] fc7 -> fc7
I0115 21:46:40.873172  7366 net.cpp:147] Setting up fc7
I0115 21:46:40.873201  7366 net.cpp:154] Top shape: 1 4096 1 1 (4096)
I0115 21:46:40.873205  7366 net.cpp:162] Memory required for data: 174878720
I0115 21:46:40.873212  7366 layer_factory.hpp:77] Creating layer bnfc7
I0115 21:46:40.873220  7366 net.cpp:103] Creating Layer bnfc7
I0115 21:46:40.873224  7366 net.cpp:451] bnfc7 <- fc7
I0115 21:46:40.873229  7366 net.cpp:394] bnfc7 -> fc7 (in-place)
I0115 21:46:40.873354  7366 net.cpp:147] Setting up bnfc7
I0115 21:46:40.873359  7366 net.cpp:154] Top shape: 1 4096 1 1 (4096)
I0115 21:46:40.873363  7366 net.cpp:162] Memory required for data: 174895104
I0115 21:46:40.873366  7366 layer_factory.hpp:77] Creating layer relu7
I0115 21:46:40.873370  7366 net.cpp:103] Creating Layer relu7
I0115 21:46:40.873373  7366 net.cpp:451] relu7 <- fc7
I0115 21:46:40.873375  7366 net.cpp:394] relu7 -> fc7 (in-place)
I0115 21:46:40.873379  7366 net.cpp:147] Setting up relu7
I0115 21:46:40.873383  7366 net.cpp:154] Top shape: 1 4096 1 1 (4096)
I0115 21:46:40.873384  7366 net.cpp:162] Memory required for data: 174911488
I0115 21:46:40.873386  7366 layer_factory.hpp:77] Creating layer fc6-deconv
I0115 21:46:40.873392  7366 net.cpp:103] Creating Layer fc6-deconv
I0115 21:46:40.873396  7366 net.cpp:451] fc6-deconv <- fc7
I0115 21:46:40.873402  7366 net.cpp:408] fc6-deconv -> fc6-deconv
I0115 21:46:42.194905  7366 net.cpp:147] Setting up fc6-deconv
I0115 21:46:42.194928  7366 net.cpp:154] Top shape: 1 512 7 7 (25088)
I0115 21:46:42.194931  7366 net.cpp:162] Memory required for data: 175011840
I0115 21:46:42.194938  7366 layer_factory.hpp:77] Creating layer fc6-deconv-bn
I0115 21:46:42.194945  7366 net.cpp:103] Creating Layer fc6-deconv-bn
I0115 21:46:42.194948  7366 net.cpp:451] fc6-deconv-bn <- fc6-deconv
I0115 21:46:42.194954  7366 net.cpp:394] fc6-deconv-bn -> fc6-deconv (in-place)
I0115 21:46:42.195124  7366 net.cpp:147] Setting up fc6-deconv-bn
I0115 21:46:42.195130  7366 net.cpp:154] Top shape: 1 512 7 7 (25088)
I0115 21:46:42.195132  7366 net.cpp:162] Memory required for data: 175112192
I0115 21:46:42.195137  7366 layer_factory.hpp:77] Creating layer fc6-deconv-relu
I0115 21:46:42.195139  7366 net.cpp:103] Creating Layer fc6-deconv-relu
I0115 21:46:42.195142  7366 net.cpp:451] fc6-deconv-relu <- fc6-deconv
I0115 21:46:42.195147  7366 net.cpp:394] fc6-deconv-relu -> fc6-deconv (in-place)
I0115 21:46:42.195150  7366 net.cpp:147] Setting up fc6-deconv-relu
I0115 21:46:42.195153  7366 net.cpp:154] Top shape: 1 512 7 7 (25088)
I0115 21:46:42.195155  7366 net.cpp:162] Memory required for data: 175212544
I0115 21:46:42.195158  7366 layer_factory.hpp:77] Creating layer unpool5
I0115 21:46:42.195161  7366 net.cpp:103] Creating Layer unpool5
I0115 21:46:42.195163  7366 net.cpp:451] unpool5 <- fc6-deconv
I0115 21:46:42.195168  7366 net.cpp:451] unpool5 <- pool5_mask
I0115 21:46:42.195173  7366 net.cpp:408] unpool5 -> unpool5
I0115 21:46:42.195185  7366 net.cpp:147] Setting up unpool5
I0115 21:46:42.195211  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:42.195214  7366 net.cpp:162] Memory required for data: 175613952
I0115 21:46:42.195217  7366 layer_factory.hpp:77] Creating layer deconv5_1
I0115 21:46:42.195225  7366 net.cpp:103] Creating Layer deconv5_1
I0115 21:46:42.195228  7366 net.cpp:451] deconv5_1 <- unpool5
I0115 21:46:42.195233  7366 net.cpp:408] deconv5_1 -> deconv5_1
I0115 21:46:42.211490  7366 net.cpp:147] Setting up deconv5_1
I0115 21:46:42.211511  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:42.211514  7366 net.cpp:162] Memory required for data: 176015360
I0115 21:46:42.211529  7366 layer_factory.hpp:77] Creating layer debn5_1
I0115 21:46:42.211536  7366 net.cpp:103] Creating Layer debn5_1
I0115 21:46:42.211540  7366 net.cpp:451] debn5_1 <- deconv5_1
I0115 21:46:42.211545  7366 net.cpp:394] debn5_1 -> deconv5_1 (in-place)
I0115 21:46:42.211714  7366 net.cpp:147] Setting up debn5_1
I0115 21:46:42.211719  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:42.211721  7366 net.cpp:162] Memory required for data: 176416768
I0115 21:46:42.211725  7366 layer_factory.hpp:77] Creating layer derelu5_1
I0115 21:46:42.211730  7366 net.cpp:103] Creating Layer derelu5_1
I0115 21:46:42.211732  7366 net.cpp:451] derelu5_1 <- deconv5_1
I0115 21:46:42.211735  7366 net.cpp:394] derelu5_1 -> deconv5_1 (in-place)
I0115 21:46:42.211740  7366 net.cpp:147] Setting up derelu5_1
I0115 21:46:42.211742  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:42.211745  7366 net.cpp:162] Memory required for data: 176818176
I0115 21:46:42.211746  7366 layer_factory.hpp:77] Creating layer deconv5_2
I0115 21:46:42.211750  7366 net.cpp:103] Creating Layer deconv5_2
I0115 21:46:42.211753  7366 net.cpp:451] deconv5_2 <- deconv5_1
I0115 21:46:42.211760  7366 net.cpp:408] deconv5_2 -> deconv5_2
I0115 21:46:42.228296  7366 net.cpp:147] Setting up deconv5_2
I0115 21:46:42.228318  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:42.228322  7366 net.cpp:162] Memory required for data: 177219584
I0115 21:46:42.228327  7366 layer_factory.hpp:77] Creating layer debn5_2
I0115 21:46:42.228355  7366 net.cpp:103] Creating Layer debn5_2
I0115 21:46:42.228375  7366 net.cpp:451] debn5_2 <- deconv5_2
I0115 21:46:42.228380  7366 net.cpp:394] debn5_2 -> deconv5_2 (in-place)
I0115 21:46:42.228549  7366 net.cpp:147] Setting up debn5_2
I0115 21:46:42.228555  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:42.228557  7366 net.cpp:162] Memory required for data: 177620992
I0115 21:46:42.228561  7366 layer_factory.hpp:77] Creating layer derelu5_2
I0115 21:46:42.228565  7366 net.cpp:103] Creating Layer derelu5_2
I0115 21:46:42.228567  7366 net.cpp:451] derelu5_2 <- deconv5_2
I0115 21:46:42.228570  7366 net.cpp:394] derelu5_2 -> deconv5_2 (in-place)
I0115 21:46:42.228574  7366 net.cpp:147] Setting up derelu5_2
I0115 21:46:42.228577  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:42.228579  7366 net.cpp:162] Memory required for data: 178022400
I0115 21:46:42.228581  7366 layer_factory.hpp:77] Creating layer deconv5_3
I0115 21:46:42.228588  7366 net.cpp:103] Creating Layer deconv5_3
I0115 21:46:42.228591  7366 net.cpp:451] deconv5_3 <- deconv5_2
I0115 21:46:42.228597  7366 net.cpp:408] deconv5_3 -> deconv5_3
I0115 21:46:42.244365  7366 net.cpp:147] Setting up deconv5_3
I0115 21:46:42.244385  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:42.244387  7366 net.cpp:162] Memory required for data: 178423808
I0115 21:46:42.244392  7366 layer_factory.hpp:77] Creating layer debn5_3
I0115 21:46:42.244400  7366 net.cpp:103] Creating Layer debn5_3
I0115 21:46:42.244403  7366 net.cpp:451] debn5_3 <- deconv5_3
I0115 21:46:42.244407  7366 net.cpp:394] debn5_3 -> deconv5_3 (in-place)
I0115 21:46:42.244532  7366 net.cpp:147] Setting up debn5_3
I0115 21:46:42.244537  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:42.244540  7366 net.cpp:162] Memory required for data: 178825216
I0115 21:46:42.244544  7366 layer_factory.hpp:77] Creating layer derelu5_3
I0115 21:46:42.244547  7366 net.cpp:103] Creating Layer derelu5_3
I0115 21:46:42.244550  7366 net.cpp:451] derelu5_3 <- deconv5_3
I0115 21:46:42.244552  7366 net.cpp:394] derelu5_3 -> deconv5_3 (in-place)
I0115 21:46:42.244556  7366 net.cpp:147] Setting up derelu5_3
I0115 21:46:42.244560  7366 net.cpp:154] Top shape: 1 512 14 14 (100352)
I0115 21:46:42.244561  7366 net.cpp:162] Memory required for data: 179226624
I0115 21:46:42.244588  7366 layer_factory.hpp:77] Creating layer unpool4
I0115 21:46:42.244593  7366 net.cpp:103] Creating Layer unpool4
I0115 21:46:42.244596  7366 net.cpp:451] unpool4 <- deconv5_3
I0115 21:46:42.244601  7366 net.cpp:451] unpool4 <- pool4_mask
I0115 21:46:42.244606  7366 net.cpp:408] unpool4 -> unpool4
I0115 21:46:42.244621  7366 net.cpp:147] Setting up unpool4
I0115 21:46:42.244626  7366 net.cpp:154] Top shape: 1 512 28 28 (401408)
I0115 21:46:42.244629  7366 net.cpp:162] Memory required for data: 180832256
I0115 21:46:42.244632  7366 layer_factory.hpp:77] Creating layer deconv4_1
I0115 21:46:42.244643  7366 net.cpp:103] Creating Layer deconv4_1
I0115 21:46:42.244647  7366 net.cpp:451] deconv4_1 <- unpool4
I0115 21:46:42.244652  7366 net.cpp:408] deconv4_1 -> deconv4_1
I0115 21:46:42.260913  7366 net.cpp:147] Setting up deconv4_1
I0115 21:46:42.260953  7366 net.cpp:154] Top shape: 1 512 28 28 (401408)
I0115 21:46:42.260957  7366 net.cpp:162] Memory required for data: 182437888
I0115 21:46:42.260965  7366 layer_factory.hpp:77] Creating layer debn4_1
I0115 21:46:42.260972  7366 net.cpp:103] Creating Layer debn4_1
I0115 21:46:42.260977  7366 net.cpp:451] debn4_1 <- deconv4_1
I0115 21:46:42.260983  7366 net.cpp:394] debn4_1 -> deconv4_1 (in-place)
I0115 21:46:42.261128  7366 net.cpp:147] Setting up debn4_1
I0115 21:46:42.261133  7366 net.cpp:154] Top shape: 1 512 28 28 (401408)
I0115 21:46:42.261137  7366 net.cpp:162] Memory required for data: 184043520
I0115 21:46:42.261139  7366 layer_factory.hpp:77] Creating layer derelu4_1
I0115 21:46:42.261143  7366 net.cpp:103] Creating Layer derelu4_1
I0115 21:46:42.261147  7366 net.cpp:451] derelu4_1 <- deconv4_1
I0115 21:46:42.261168  7366 net.cpp:394] derelu4_1 -> deconv4_1 (in-place)
I0115 21:46:42.261173  7366 net.cpp:147] Setting up derelu4_1
I0115 21:46:42.261178  7366 net.cpp:154] Top shape: 1 512 28 28 (401408)
I0115 21:46:42.261180  7366 net.cpp:162] Memory required for data: 185649152
I0115 21:46:42.261183  7366 layer_factory.hpp:77] Creating layer deconv4_2
I0115 21:46:42.261188  7366 net.cpp:103] Creating Layer deconv4_2
I0115 21:46:42.261190  7366 net.cpp:451] deconv4_2 <- deconv4_1
I0115 21:46:42.261194  7366 net.cpp:408] deconv4_2 -> deconv4_2
I0115 21:46:42.277176  7366 net.cpp:147] Setting up deconv4_2
I0115 21:46:42.277217  7366 net.cpp:154] Top shape: 1 512 28 28 (401408)
I0115 21:46:42.277220  7366 net.cpp:162] Memory required for data: 187254784
I0115 21:46:42.277228  7366 layer_factory.hpp:77] Creating layer debn4_2
I0115 21:46:42.277237  7366 net.cpp:103] Creating Layer debn4_2
I0115 21:46:42.277241  7366 net.cpp:451] debn4_2 <- deconv4_2
I0115 21:46:42.277248  7366 net.cpp:394] debn4_2 -> deconv4_2 (in-place)
I0115 21:46:42.277388  7366 net.cpp:147] Setting up debn4_2
I0115 21:46:42.277395  7366 net.cpp:154] Top shape: 1 512 28 28 (401408)
I0115 21:46:42.277396  7366 net.cpp:162] Memory required for data: 188860416
I0115 21:46:42.277400  7366 layer_factory.hpp:77] Creating layer derelu4_2
I0115 21:46:42.277405  7366 net.cpp:103] Creating Layer derelu4_2
I0115 21:46:42.277406  7366 net.cpp:451] derelu4_2 <- deconv4_2
I0115 21:46:42.277427  7366 net.cpp:394] derelu4_2 -> deconv4_2 (in-place)
I0115 21:46:42.277431  7366 net.cpp:147] Setting up derelu4_2
I0115 21:46:42.277436  7366 net.cpp:154] Top shape: 1 512 28 28 (401408)
I0115 21:46:42.277438  7366 net.cpp:162] Memory required for data: 190466048
I0115 21:46:42.277441  7366 layer_factory.hpp:77] Creating layer deconv4_3
I0115 21:46:42.277446  7366 net.cpp:103] Creating Layer deconv4_3
I0115 21:46:42.277449  7366 net.cpp:451] deconv4_3 <- deconv4_2
I0115 21:46:42.277454  7366 net.cpp:408] deconv4_3 -> deconv4_3
I0115 21:46:42.285833  7366 net.cpp:147] Setting up deconv4_3
I0115 21:46:42.285851  7366 net.cpp:154] Top shape: 1 256 28 28 (200704)
I0115 21:46:42.285853  7366 net.cpp:162] Memory required for data: 191268864
I0115 21:46:42.285858  7366 layer_factory.hpp:77] Creating layer debn4_3
I0115 21:46:42.285882  7366 net.cpp:103] Creating Layer debn4_3
I0115 21:46:42.285887  7366 net.cpp:451] debn4_3 <- deconv4_3
I0115 21:46:42.285894  7366 net.cpp:394] debn4_3 -> deconv4_3 (in-place)
I0115 21:46:42.286037  7366 net.cpp:147] Setting up debn4_3
I0115 21:46:42.286043  7366 net.cpp:154] Top shape: 1 256 28 28 (200704)
I0115 21:46:42.286064  7366 net.cpp:162] Memory required for data: 192071680
I0115 21:46:42.286068  7366 layer_factory.hpp:77] Creating layer derelu4_3
I0115 21:46:42.286072  7366 net.cpp:103] Creating Layer derelu4_3
I0115 21:46:42.286075  7366 net.cpp:451] derelu4_3 <- deconv4_3
I0115 21:46:42.286078  7366 net.cpp:394] derelu4_3 -> deconv4_3 (in-place)
I0115 21:46:42.286083  7366 net.cpp:147] Setting up derelu4_3
I0115 21:46:42.286087  7366 net.cpp:154] Top shape: 1 256 28 28 (200704)
I0115 21:46:42.286088  7366 net.cpp:162] Memory required for data: 192874496
I0115 21:46:42.286092  7366 layer_factory.hpp:77] Creating layer unpool3
I0115 21:46:42.286098  7366 net.cpp:103] Creating Layer unpool3
I0115 21:46:42.286099  7366 net.cpp:451] unpool3 <- deconv4_3
I0115 21:46:42.286103  7366 net.cpp:451] unpool3 <- pool3_mask
I0115 21:46:42.286106  7366 net.cpp:408] unpool3 -> unpool3
I0115 21:46:42.286121  7366 net.cpp:147] Setting up unpool3
I0115 21:46:42.286125  7366 net.cpp:154] Top shape: 1 256 56 56 (802816)
I0115 21:46:42.286128  7366 net.cpp:162] Memory required for data: 196085760
I0115 21:46:42.286129  7366 layer_factory.hpp:77] Creating layer deconv3_1
I0115 21:46:42.286135  7366 net.cpp:103] Creating Layer deconv3_1
I0115 21:46:42.286139  7366 net.cpp:451] deconv3_1 <- unpool3
I0115 21:46:42.286144  7366 net.cpp:408] deconv3_1 -> deconv3_1
I0115 21:46:42.290585  7366 net.cpp:147] Setting up deconv3_1
I0115 21:46:42.290599  7366 net.cpp:154] Top shape: 1 256 56 56 (802816)
I0115 21:46:42.290602  7366 net.cpp:162] Memory required for data: 199297024
I0115 21:46:42.290607  7366 layer_factory.hpp:77] Creating layer debn3_1
I0115 21:46:42.290613  7366 net.cpp:103] Creating Layer debn3_1
I0115 21:46:42.290616  7366 net.cpp:451] debn3_1 <- deconv3_1
I0115 21:46:42.290621  7366 net.cpp:394] debn3_1 -> deconv3_1 (in-place)
I0115 21:46:42.290783  7366 net.cpp:147] Setting up debn3_1
I0115 21:46:42.290789  7366 net.cpp:154] Top shape: 1 256 56 56 (802816)
I0115 21:46:42.290791  7366 net.cpp:162] Memory required for data: 202508288
I0115 21:46:42.290796  7366 layer_factory.hpp:77] Creating layer derelu3_1
I0115 21:46:42.290798  7366 net.cpp:103] Creating Layer derelu3_1
I0115 21:46:42.290802  7366 net.cpp:451] derelu3_1 <- deconv3_1
I0115 21:46:42.290805  7366 net.cpp:394] derelu3_1 -> deconv3_1 (in-place)
I0115 21:46:42.290808  7366 net.cpp:147] Setting up derelu3_1
I0115 21:46:42.290812  7366 net.cpp:154] Top shape: 1 256 56 56 (802816)
I0115 21:46:42.290814  7366 net.cpp:162] Memory required for data: 205719552
I0115 21:46:42.290817  7366 layer_factory.hpp:77] Creating layer deconv3_2
I0115 21:46:42.290822  7366 net.cpp:103] Creating Layer deconv3_2
I0115 21:46:42.290824  7366 net.cpp:451] deconv3_2 <- deconv3_1
I0115 21:46:42.290828  7366 net.cpp:408] deconv3_2 -> deconv3_2
I0115 21:46:42.295343  7366 net.cpp:147] Setting up deconv3_2
I0115 21:46:42.295374  7366 net.cpp:154] Top shape: 1 256 56 56 (802816)
I0115 21:46:42.295377  7366 net.cpp:162] Memory required for data: 208930816
I0115 21:46:42.295383  7366 layer_factory.hpp:77] Creating layer debn3_2
I0115 21:46:42.295390  7366 net.cpp:103] Creating Layer debn3_2
I0115 21:46:42.295394  7366 net.cpp:451] debn3_2 <- deconv3_2
I0115 21:46:42.295399  7366 net.cpp:394] debn3_2 -> deconv3_2 (in-place)
I0115 21:46:42.295542  7366 net.cpp:147] Setting up debn3_2
I0115 21:46:42.295548  7366 net.cpp:154] Top shape: 1 256 56 56 (802816)
I0115 21:46:42.295550  7366 net.cpp:162] Memory required for data: 212142080
I0115 21:46:42.295554  7366 layer_factory.hpp:77] Creating layer derelu3_2
I0115 21:46:42.295578  7366 net.cpp:103] Creating Layer derelu3_2
I0115 21:46:42.295581  7366 net.cpp:451] derelu3_2 <- deconv3_2
I0115 21:46:42.295585  7366 net.cpp:394] derelu3_2 -> deconv3_2 (in-place)
I0115 21:46:42.295590  7366 net.cpp:147] Setting up derelu3_2
I0115 21:46:42.295594  7366 net.cpp:154] Top shape: 1 256 56 56 (802816)
I0115 21:46:42.295598  7366 net.cpp:162] Memory required for data: 215353344
I0115 21:46:42.295599  7366 layer_factory.hpp:77] Creating layer deconv3_3
I0115 21:46:42.295604  7366 net.cpp:103] Creating Layer deconv3_3
I0115 21:46:42.295608  7366 net.cpp:451] deconv3_3 <- deconv3_2
I0115 21:46:42.295614  7366 net.cpp:408] deconv3_3 -> deconv3_3
I0115 21:46:42.298081  7366 net.cpp:147] Setting up deconv3_3
I0115 21:46:42.298090  7366 net.cpp:154] Top shape: 1 128 56 56 (401408)
I0115 21:46:42.298112  7366 net.cpp:162] Memory required for data: 216958976
I0115 21:46:42.298117  7366 layer_factory.hpp:77] Creating layer debn3_3
I0115 21:46:42.298123  7366 net.cpp:103] Creating Layer debn3_3
I0115 21:46:42.298127  7366 net.cpp:451] debn3_3 <- deconv3_3
I0115 21:46:42.298131  7366 net.cpp:394] debn3_3 -> deconv3_3 (in-place)
I0115 21:46:42.298274  7366 net.cpp:147] Setting up debn3_3
I0115 21:46:42.298280  7366 net.cpp:154] Top shape: 1 128 56 56 (401408)
I0115 21:46:42.298281  7366 net.cpp:162] Memory required for data: 218564608
I0115 21:46:42.298285  7366 layer_factory.hpp:77] Creating layer derelu3_3
I0115 21:46:42.298306  7366 net.cpp:103] Creating Layer derelu3_3
I0115 21:46:42.298310  7366 net.cpp:451] derelu3_3 <- deconv3_3
I0115 21:46:42.298313  7366 net.cpp:394] derelu3_3 -> deconv3_3 (in-place)
I0115 21:46:42.298317  7366 net.cpp:147] Setting up derelu3_3
I0115 21:46:42.298321  7366 net.cpp:154] Top shape: 1 128 56 56 (401408)
I0115 21:46:42.298323  7366 net.cpp:162] Memory required for data: 220170240
I0115 21:46:42.298326  7366 layer_factory.hpp:77] Creating layer unpool2
I0115 21:46:42.298332  7366 net.cpp:103] Creating Layer unpool2
I0115 21:46:42.298336  7366 net.cpp:451] unpool2 <- deconv3_3
I0115 21:46:42.298338  7366 net.cpp:451] unpool2 <- pool2_mask
I0115 21:46:42.298343  7366 net.cpp:408] unpool2 -> unpool2
I0115 21:46:42.298359  7366 net.cpp:147] Setting up unpool2
I0115 21:46:42.298364  7366 net.cpp:154] Top shape: 1 128 112 112 (1605632)
I0115 21:46:42.298367  7366 net.cpp:162] Memory required for data: 226592768
I0115 21:46:42.298370  7366 layer_factory.hpp:77] Creating layer deconv2_1
I0115 21:46:42.298377  7366 net.cpp:103] Creating Layer deconv2_1
I0115 21:46:42.298379  7366 net.cpp:451] deconv2_1 <- unpool2
I0115 21:46:42.298384  7366 net.cpp:408] deconv2_1 -> deconv2_1
I0115 21:46:42.299388  7366 net.cpp:147] Setting up deconv2_1
I0115 21:46:42.299394  7366 net.cpp:154] Top shape: 1 128 112 112 (1605632)
I0115 21:46:42.299397  7366 net.cpp:162] Memory required for data: 233015296
I0115 21:46:42.299402  7366 layer_factory.hpp:77] Creating layer debn2_1
I0115 21:46:42.299408  7366 net.cpp:103] Creating Layer debn2_1
I0115 21:46:42.299412  7366 net.cpp:451] debn2_1 <- deconv2_1
I0115 21:46:42.299415  7366 net.cpp:394] debn2_1 -> deconv2_1 (in-place)
I0115 21:46:42.299544  7366 net.cpp:147] Setting up debn2_1
I0115 21:46:42.299551  7366 net.cpp:154] Top shape: 1 128 112 112 (1605632)
I0115 21:46:42.299553  7366 net.cpp:162] Memory required for data: 239437824
I0115 21:46:42.299557  7366 layer_factory.hpp:77] Creating layer derelu2_1
I0115 21:46:42.299561  7366 net.cpp:103] Creating Layer derelu2_1
I0115 21:46:42.299566  7366 net.cpp:451] derelu2_1 <- deconv2_1
I0115 21:46:42.299568  7366 net.cpp:394] derelu2_1 -> deconv2_1 (in-place)
I0115 21:46:42.299574  7366 net.cpp:147] Setting up derelu2_1
I0115 21:46:42.299578  7366 net.cpp:154] Top shape: 1 128 112 112 (1605632)
I0115 21:46:42.299582  7366 net.cpp:162] Memory required for data: 245860352
I0115 21:46:42.299583  7366 layer_factory.hpp:77] Creating layer deconv2_2
I0115 21:46:42.299588  7366 net.cpp:103] Creating Layer deconv2_2
I0115 21:46:42.299592  7366 net.cpp:451] deconv2_2 <- deconv2_1
I0115 21:46:42.299595  7366 net.cpp:408] deconv2_2 -> deconv2_2
I0115 21:46:42.300174  7366 net.cpp:147] Setting up deconv2_2
I0115 21:46:42.300180  7366 net.cpp:154] Top shape: 1 64 112 112 (802816)
I0115 21:46:42.300184  7366 net.cpp:162] Memory required for data: 249071616
I0115 21:46:42.300189  7366 layer_factory.hpp:77] Creating layer debn2_2
I0115 21:46:42.300194  7366 net.cpp:103] Creating Layer debn2_2
I0115 21:46:42.300197  7366 net.cpp:451] debn2_2 <- deconv2_2
I0115 21:46:42.300201  7366 net.cpp:394] debn2_2 -> deconv2_2 (in-place)
I0115 21:46:42.300372  7366 net.cpp:147] Setting up debn2_2
I0115 21:46:42.300379  7366 net.cpp:154] Top shape: 1 64 112 112 (802816)
I0115 21:46:42.300381  7366 net.cpp:162] Memory required for data: 252282880
I0115 21:46:42.300386  7366 layer_factory.hpp:77] Creating layer derelu2_2
I0115 21:46:42.300390  7366 net.cpp:103] Creating Layer derelu2_2
I0115 21:46:42.300393  7366 net.cpp:451] derelu2_2 <- deconv2_2
I0115 21:46:42.300396  7366 net.cpp:394] derelu2_2 -> deconv2_2 (in-place)
I0115 21:46:42.300400  7366 net.cpp:147] Setting up derelu2_2
I0115 21:46:42.300405  7366 net.cpp:154] Top shape: 1 64 112 112 (802816)
I0115 21:46:42.300407  7366 net.cpp:162] Memory required for data: 255494144
I0115 21:46:42.300410  7366 layer_factory.hpp:77] Creating layer unpool1
I0115 21:46:42.300415  7366 net.cpp:103] Creating Layer unpool1
I0115 21:46:42.300417  7366 net.cpp:451] unpool1 <- deconv2_2
I0115 21:46:42.300421  7366 net.cpp:451] unpool1 <- pool1_mask
I0115 21:46:42.300426  7366 net.cpp:408] unpool1 -> unpool1
I0115 21:46:42.300441  7366 net.cpp:147] Setting up unpool1
I0115 21:46:42.300446  7366 net.cpp:154] Top shape: 1 64 224 224 (3211264)
I0115 21:46:42.300449  7366 net.cpp:162] Memory required for data: 268339200
I0115 21:46:42.300452  7366 layer_factory.hpp:77] Creating layer deconv1_1
I0115 21:46:42.300457  7366 net.cpp:103] Creating Layer deconv1_1
I0115 21:46:42.300460  7366 net.cpp:451] deconv1_1 <- unpool1
I0115 21:46:42.300465  7366 net.cpp:408] deconv1_1 -> deconv1_1
I0115 21:46:42.300839  7366 net.cpp:147] Setting up deconv1_1
I0115 21:46:42.300845  7366 net.cpp:154] Top shape: 1 64 224 224 (3211264)
I0115 21:46:42.300848  7366 net.cpp:162] Memory required for data: 281184256
I0115 21:46:42.300853  7366 layer_factory.hpp:77] Creating layer debn1_1
I0115 21:46:42.300858  7366 net.cpp:103] Creating Layer debn1_1
I0115 21:46:42.300861  7366 net.cpp:451] debn1_1 <- deconv1_1
I0115 21:46:42.300865  7366 net.cpp:394] debn1_1 -> deconv1_1 (in-place)
I0115 21:46:42.301594  7366 net.cpp:147] Setting up debn1_1
I0115 21:46:42.301602  7366 net.cpp:154] Top shape: 1 64 224 224 (3211264)
I0115 21:46:42.301605  7366 net.cpp:162] Memory required for data: 294029312
I0115 21:46:42.301610  7366 layer_factory.hpp:77] Creating layer derelu1_1
I0115 21:46:42.301615  7366 net.cpp:103] Creating Layer derelu1_1
I0115 21:46:42.301618  7366 net.cpp:451] derelu1_1 <- deconv1_1
I0115 21:46:42.301622  7366 net.cpp:394] derelu1_1 -> deconv1_1 (in-place)
I0115 21:46:42.301627  7366 net.cpp:147] Setting up derelu1_1
I0115 21:46:42.301631  7366 net.cpp:154] Top shape: 1 64 224 224 (3211264)
I0115 21:46:42.301633  7366 net.cpp:162] Memory required for data: 306874368
I0115 21:46:42.301636  7366 layer_factory.hpp:77] Creating layer deconv1_2
I0115 21:46:42.301643  7366 net.cpp:103] Creating Layer deconv1_2
I0115 21:46:42.301646  7366 net.cpp:451] deconv1_2 <- deconv1_1
I0115 21:46:42.301651  7366 net.cpp:408] deconv1_2 -> deconv1_2
I0115 21:46:42.302021  7366 net.cpp:147] Setting up deconv1_2
I0115 21:46:42.302027  7366 net.cpp:154] Top shape: 1 64 224 224 (3211264)
I0115 21:46:42.302031  7366 net.cpp:162] Memory required for data: 319719424
I0115 21:46:42.302034  7366 layer_factory.hpp:77] Creating layer debn1_2
I0115 21:46:42.302040  7366 net.cpp:103] Creating Layer debn1_2
I0115 21:46:42.302043  7366 net.cpp:451] debn1_2 <- deconv1_2
I0115 21:46:42.302047  7366 net.cpp:394] debn1_2 -> deconv1_2 (in-place)
I0115 21:46:42.302187  7366 net.cpp:147] Setting up debn1_2
I0115 21:46:42.302193  7366 net.cpp:154] Top shape: 1 64 224 224 (3211264)
I0115 21:46:42.302196  7366 net.cpp:162] Memory required for data: 332564480
I0115 21:46:42.302201  7366 layer_factory.hpp:77] Creating layer derelu1_2
I0115 21:46:42.302204  7366 net.cpp:103] Creating Layer derelu1_2
I0115 21:46:42.302208  7366 net.cpp:451] derelu1_2 <- deconv1_2
I0115 21:46:42.302211  7366 net.cpp:394] derelu1_2 -> deconv1_2 (in-place)
I0115 21:46:42.302215  7366 net.cpp:147] Setting up derelu1_2
I0115 21:46:42.302219  7366 net.cpp:154] Top shape: 1 64 224 224 (3211264)
I0115 21:46:42.302222  7366 net.cpp:162] Memory required for data: 345409536
I0115 21:46:42.302225  7366 layer_factory.hpp:77] Creating layer class_score_nyu
I0115 21:46:42.302232  7366 net.cpp:103] Creating Layer class_score_nyu
I0115 21:46:42.302235  7366 net.cpp:451] class_score_nyu <- deconv1_2
I0115 21:46:42.302240  7366 net.cpp:408] class_score_nyu -> class_score
I0115 21:46:42.302402  7366 net.cpp:147] Setting up class_score_nyu
I0115 21:46:42.302408  7366 net.cpp:154] Top shape: 1 14 224 224 (702464)
I0115 21:46:42.302412  7366 net.cpp:162] Memory required for data: 348219392
I0115 21:46:42.302417  7366 layer_factory.hpp:77] Creating layer probability
I0115 21:46:42.302420  7366 net.cpp:103] Creating Layer probability
I0115 21:46:42.302424  7366 net.cpp:451] probability <- class_score
I0115 21:46:42.302429  7366 net.cpp:408] probability -> probability
I0115 21:46:42.302474  7366 net.cpp:147] Setting up probability
I0115 21:46:42.302479  7366 net.cpp:154] Top shape: 1 14 224 224 (702464)
I0115 21:46:42.302482  7366 net.cpp:162] Memory required for data: 351029248
I0115 21:46:42.302485  7366 net.cpp:225] probability does not need backward computation.
I0115 21:46:42.302496  7366 net.cpp:225] class_score_nyu does not need backward computation.
I0115 21:46:42.302500  7366 net.cpp:225] derelu1_2 does not need backward computation.
I0115 21:46:42.302501  7366 net.cpp:225] debn1_2 does not need backward computation.
I0115 21:46:42.302505  7366 net.cpp:225] deconv1_2 does not need backward computation.
I0115 21:46:42.302508  7366 net.cpp:225] derelu1_1 does not need backward computation.
I0115 21:46:42.302511  7366 net.cpp:225] debn1_1 does not need backward computation.
I0115 21:46:42.302513  7366 net.cpp:225] deconv1_1 does not need backward computation.
I0115 21:46:42.302516  7366 net.cpp:225] unpool1 does not need backward computation.
I0115 21:46:42.302520  7366 net.cpp:225] derelu2_2 does not need backward computation.
I0115 21:46:42.302522  7366 net.cpp:225] debn2_2 does not need backward computation.
I0115 21:46:42.302525  7366 net.cpp:225] deconv2_2 does not need backward computation.
I0115 21:46:42.302528  7366 net.cpp:225] derelu2_1 does not need backward computation.
I0115 21:46:42.302531  7366 net.cpp:225] debn2_1 does not need backward computation.
I0115 21:46:42.302534  7366 net.cpp:225] deconv2_1 does not need backward computation.
I0115 21:46:42.302537  7366 net.cpp:225] unpool2 does not need backward computation.
I0115 21:46:42.302541  7366 net.cpp:225] derelu3_3 does not need backward computation.
I0115 21:46:42.302543  7366 net.cpp:225] debn3_3 does not need backward computation.
I0115 21:46:42.302546  7366 net.cpp:225] deconv3_3 does not need backward computation.
I0115 21:46:42.302548  7366 net.cpp:225] derelu3_2 does not need backward computation.
I0115 21:46:42.302551  7366 net.cpp:225] debn3_2 does not need backward computation.
I0115 21:46:42.302554  7366 net.cpp:225] deconv3_2 does not need backward computation.
I0115 21:46:42.302557  7366 net.cpp:225] derelu3_1 does not need backward computation.
I0115 21:46:42.302561  7366 net.cpp:225] debn3_1 does not need backward computation.
I0115 21:46:42.302563  7366 net.cpp:225] deconv3_1 does not need backward computation.
I0115 21:46:42.302567  7366 net.cpp:225] unpool3 does not need backward computation.
I0115 21:46:42.302570  7366 net.cpp:225] derelu4_3 does not need backward computation.
I0115 21:46:42.302573  7366 net.cpp:225] debn4_3 does not need backward computation.
I0115 21:46:42.302577  7366 net.cpp:225] deconv4_3 does not need backward computation.
I0115 21:46:42.302579  7366 net.cpp:225] derelu4_2 does not need backward computation.
I0115 21:46:42.302582  7366 net.cpp:225] debn4_2 does not need backward computation.
I0115 21:46:42.302585  7366 net.cpp:225] deconv4_2 does not need backward computation.
I0115 21:46:42.302588  7366 net.cpp:225] derelu4_1 does not need backward computation.
I0115 21:46:42.302592  7366 net.cpp:225] debn4_1 does not need backward computation.
I0115 21:46:42.302594  7366 net.cpp:225] deconv4_1 does not need backward computation.
I0115 21:46:42.302598  7366 net.cpp:225] unpool4 does not need backward computation.
I0115 21:46:42.302603  7366 net.cpp:225] derelu5_3 does not need backward computation.
I0115 21:46:42.302605  7366 net.cpp:225] debn5_3 does not need backward computation.
I0115 21:46:42.302609  7366 net.cpp:225] deconv5_3 does not need backward computation.
I0115 21:46:42.302613  7366 net.cpp:225] derelu5_2 does not need backward computation.
I0115 21:46:42.302615  7366 net.cpp:225] debn5_2 does not need backward computation.
I0115 21:46:42.302618  7366 net.cpp:225] deconv5_2 does not need backward computation.
I0115 21:46:42.302621  7366 net.cpp:225] derelu5_1 does not need backward computation.
I0115 21:46:42.302624  7366 net.cpp:225] debn5_1 does not need backward computation.
I0115 21:46:42.302628  7366 net.cpp:225] deconv5_1 does not need backward computation.
I0115 21:46:42.302630  7366 net.cpp:225] unpool5 does not need backward computation.
I0115 21:46:42.302634  7366 net.cpp:225] fc6-deconv-relu does not need backward computation.
I0115 21:46:42.302639  7366 net.cpp:225] fc6-deconv-bn does not need backward computation.
I0115 21:46:42.302641  7366 net.cpp:225] fc6-deconv does not need backward computation.
I0115 21:46:42.302644  7366 net.cpp:225] relu7 does not need backward computation.
I0115 21:46:42.302647  7366 net.cpp:225] bnfc7 does not need backward computation.
I0115 21:46:42.302650  7366 net.cpp:225] fc7 does not need backward computation.
I0115 21:46:42.302654  7366 net.cpp:225] relu6 does not need backward computation.
I0115 21:46:42.302657  7366 net.cpp:225] bnfc6 does not need backward computation.
I0115 21:46:42.302660  7366 net.cpp:225] fc6 does not need backward computation.
I0115 21:46:42.302664  7366 net.cpp:225] pool5 does not need backward computation.
I0115 21:46:42.302667  7366 net.cpp:225] relu5_3 does not need backward computation.
I0115 21:46:42.302670  7366 net.cpp:225] bn5_3 does not need backward computation.
I0115 21:46:42.302673  7366 net.cpp:225] conv5_3 does not need backward computation.
I0115 21:46:42.302676  7366 net.cpp:225] relu5_2 does not need backward computation.
I0115 21:46:42.302680  7366 net.cpp:225] bn5_2 does not need backward computation.
I0115 21:46:42.302682  7366 net.cpp:225] conv5_2 does not need backward computation.
I0115 21:46:42.302685  7366 net.cpp:225] relu5_1 does not need backward computation.
I0115 21:46:42.302690  7366 net.cpp:225] bn5_1 does not need backward computation.
I0115 21:46:42.302692  7366 net.cpp:225] conv5_1 does not need backward computation.
I0115 21:46:42.302695  7366 net.cpp:225] pool4 does not need backward computation.
I0115 21:46:42.302698  7366 net.cpp:225] relu4_3 does not need backward computation.
I0115 21:46:42.302701  7366 net.cpp:225] bn4_3 does not need backward computation.
I0115 21:46:42.302704  7366 net.cpp:225] conv4_3 does not need backward computation.
I0115 21:46:42.302707  7366 net.cpp:225] relu4_2 does not need backward computation.
I0115 21:46:42.302711  7366 net.cpp:225] bn4_2 does not need backward computation.
I0115 21:46:42.302713  7366 net.cpp:225] conv4_2 does not need backward computation.
I0115 21:46:42.302716  7366 net.cpp:225] relu4_1 does not need backward computation.
I0115 21:46:42.302719  7366 net.cpp:225] bn4_1 does not need backward computation.
I0115 21:46:42.302722  7366 net.cpp:225] conv4_1 does not need backward computation.
I0115 21:46:42.302726  7366 net.cpp:225] pool3 does not need backward computation.
I0115 21:46:42.302729  7366 net.cpp:225] relu3_3 does not need backward computation.
I0115 21:46:42.302732  7366 net.cpp:225] bn3_3 does not need backward computation.
I0115 21:46:42.302736  7366 net.cpp:225] conv3_3 does not need backward computation.
I0115 21:46:42.302738  7366 net.cpp:225] relu3_2 does not need backward computation.
I0115 21:46:42.302742  7366 net.cpp:225] bn3_2 does not need backward computation.
I0115 21:46:42.302744  7366 net.cpp:225] conv3_2 does not need backward computation.
I0115 21:46:42.302748  7366 net.cpp:225] relu3_1 does not need backward computation.
I0115 21:46:42.302752  7366 net.cpp:225] bn3_1 does not need backward computation.
I0115 21:46:42.302754  7366 net.cpp:225] conv3_1 does not need backward computation.
I0115 21:46:42.302757  7366 net.cpp:225] pool2 does not need backward computation.
I0115 21:46:42.302760  7366 net.cpp:225] relu2_2 does not need backward computation.
I0115 21:46:42.302763  7366 net.cpp:225] bn2_2 does not need backward computation.
I0115 21:46:42.302767  7366 net.cpp:225] conv2_2 does not need backward computation.
I0115 21:46:42.302770  7366 net.cpp:225] relu2_1 does not need backward computation.
I0115 21:46:42.302773  7366 net.cpp:225] bn2_1 does not need backward computation.
I0115 21:46:42.302776  7366 net.cpp:225] conv2_1 does not need backward computation.
I0115 21:46:42.302779  7366 net.cpp:225] pool1 does not need backward computation.
I0115 21:46:42.302783  7366 net.cpp:225] relu1_2 does not need backward computation.
I0115 21:46:42.302786  7366 net.cpp:225] bn1_2 does not need backward computation.
I0115 21:46:42.302789  7366 net.cpp:225] conv1_2 does not need backward computation.
I0115 21:46:42.302793  7366 net.cpp:225] relu1_1 does not need backward computation.
I0115 21:46:42.302795  7366 net.cpp:225] bn1_1 does not need backward computation.
I0115 21:46:42.302798  7366 net.cpp:225] conv1_1 does not need backward computation.
I0115 21:46:42.302801  7366 net.cpp:267] This network produces output probability
I0115 21:46:42.302834  7366 net.cpp:280] Network initialization done.
[libprotobuf WARNING google/protobuf/io/coded_stream.cc:537] Reading dangerously large protocol message.  If the message turns out to be larger than 2147483647 bytes, parsing will be halted for security reasons.  To increase the limit (or to disable these warnings), see CodedInputStream::SetTotalBytesLimit() in google/protobuf/io/coded_stream.h.
[libprotobuf WARNING google/protobuf/io/coded_stream.cc:78] The total number of bytes read was 1007233675
I0115 21:46:42.764060  7366 net.cpp:813] Ignoring source layer data
I0115 21:46:42.764072  7366 net.cpp:813] Ignoring source layer 
I0115 21:46:42.971236  7366 net.cpp:813] Ignoring source layer deconv1_2_derelu1_2_0_split
I0115 21:46:42.971259  7366 net.cpp:813] Ignoring source layer instance_score
Network produces 14 output classes
Creating live capture... OpenNI2-FreenectDriver: Using libfreenect v0.6.0
Skipping Kinect v2 device (needs https://github.com/libfreenect2).
Skipping Kinect v2 device (needs https://github.com/libfreenect2).
failed!
DeviceOpen using default: no devices found
Error initialising live device...

出现“DeviceOpen using default: no devices found   Error initialising live device...”问题,是因为我没有将libfreenect2/build/lib中的libfreenect2-openni2.so 和 libfreenect2-openni2.so.0放置在OpenNI2/Bin/x64-Release/OpenNI2/Drivers中。即rgbd相机驱动没有放在正确的位置。

将驱动问题解决好,再次运行,又出现如下问题:

Creating live capture... OpenNI2-FreenectDriver: Using libfreenect v0.6.0
Skipping Kinect v2 device (needs https://github.com/libfreenect2).
Skipping Kinect v2 device (needs https://github.com/libfreenect2).
[Info] [Freenect2Impl] enumerating devices...
[Info] [Freenect2Impl] 8 usb devices connected
[Info] [Freenect2Impl] found valid Kinect v2 @2:3 with serial 034011551247
[Info] [Freenect2Impl] found 1 devices
libva info: VA-API version 0.39.0
libva info: va_getDriverName() returns 0
libva info: Trying to open /usr/lib/x86_64-linux-gnu/dri/i965_drv_video.so
libva info: Found init function __vaDriverInit_0_39
libva error: /usr/lib/x86_64-linux-gnu/dri/i965_drv_video.so init failed
libva info: va_openDriver() returns -1
[Error] [VaapiRgbPacketProcessorImpl] vaInitialize(display, &major_ver, &minor_ver): unknown libva error
[Info] [Freenect2DeviceImpl] opening...
[Info] [Freenect2DeviceImpl] transfer pool sizes rgb: 20*16384 ir: 60*8*33792
[Info] [Freenect2DeviceImpl] opened
[Info] [Freenect2DeviceImpl] starting...
[Info] [Freenect2DeviceImpl] submitting rgb transfers...
[Info] [Freenect2DeviceImpl] submitting depth transfers...
[Info] [Freenect2DeviceImpl] started
success!
Waiting for first frame..............fx=1081.37,fy=1081.37,cx=959.5,cy=539.5
fx=364.318,fy=364.318,ix=260.671,iy=201.638,k1=0.092919,k2=-0.271412,k3=0.0976559,p1=0,p2=0
... got it!
[Info] [DepthPacketStreamParser] 2 packets were lost
F0212 22:54:32.332582  9590 syncedmem.cpp:56] Check failed: error == cudaSuccess (2 vs. 0)  out of memory
*** Check failure stack trace: ***
Aborted (core dumped)

运行数据库,也出现同样的问题:

./SemanticFusion /path/to/your/nyu_data_small/bathroom_0003.txt /path/to/your/nyu_data_small/output_predictions.txt
#我的路径如下:
./SemanticFusion /home/fang/Downloads/dataset/nyu_data_small/bathroom_0003.txt /home/fang/Downloads/dataset/nyu_data_small/output_predictions.txt
Looking for RGB/Depth images in folder:/home/fang/Downloads/dataset/nyu_data_small
E.g.:/home/fang/Downloads/dataset/nyu_data_small/./bathroom_0003/1_depth.png
Found:/home/fang/Downloads/dataset/nyu_data_small/./bathroom_0003/301_depth.png
Segmentation fault (core dumped)

Looking for RGB/Depth images in folder:/home/fang/Downloads/dataset/nyu_data_small
E.g.:/home/fang/Downloads/dataset/nyu_data_small/./bathroom_0003/1_depth.png
Found:/home/fang/Downloads/dataset/nyu_data_small/./bathroom_0003/301_depth.png
F0114 23:36:03.832850 17462 syncedmem.cpp:56] Check failed: error == cudaSuccess (2 vs. 0)  out of memory
*** Check failure stack trace: ***
Aborted (core dumped)

这个问题可能是由于显卡内存不够,我的笔记本显卡类型是GeForce GTX 1060 with Max-Q Design ,显存大小为3G。

  • 2
    点赞
  • 2
    收藏
    觉得还不错? 一键收藏
  • 3
    评论

“相关推荐”对你有帮助么?

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
评论 3
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值