Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[dnn] Ability to load FSSD pre trained caffe model with "Interp" layer #10985

Closed
fabito opened this issue Mar 5, 2018 · 2 comments
Closed

[dnn] Ability to load FSSD pre trained caffe model with "Interp" layer #10985

fabito opened this issue Mar 5, 2018 · 2 comments
Assignees
@dkurt
Labels
category: dnn feature

Comments

@fabito
Copy link

fabito commented Mar 5, 2018 •
edited by dkurt

I'm trying to load a FSSD: Feature Fusion Single Shot Multibox Detector pre trained caffe model.
But I'm getting the following error:

[libprotobuf ERROR /io/opencv/3rdparty/protobuf/src/google/protobuf/text_format.cc:298] Error parsing text-format opencv_caffe.NetParameter: 751:16: Message type "opencv_caffe.LayerParameter" has no field named "interp_param".
OpenCV Error: Unspecified error (FAILED: ReadProtoFromTextFile(param_file, param). Failed to parse NetParameter file: /zoo/caffe-zoo/FSSD_512_COCO/deploy.prototxt) in ReadNetParamsFromTextFileOrDie, file /io/opencv/modules/dnn/src/caffe/caffe_io.cpp, line 1145
Traceback (most recent call last):
  File "vigilante/play.py", line 86, in <module>
    prototxt='/zoo/caffe-zoo/FSSD_512_COCO/deploy.prototxt')
  File "/usr/src/app/vigilante/object_detector.py", line 192, in __init__
    self.net = cv2.dnn.readNetFromCaffe(prototxt, caffemodel)
cv2.error: /io/opencv/modules/dnn/src/caffe/caffe_io.cpp:1145: error: (-2) FAILED: ReadProtoFromTextFile(param_file, param). Failed to parse NetParameter file: /zoo/caffe-zoo/FSSD_512_COCO/deploy.prototxt in function ReadNetParamsFromTextFileOrDie
deploy.prototxt 
name: "VGG_COCO_SSD_FPN_NO33_WOP512x512_deploy"
input: "data"
input_shape {
  dim: 1
  dim: 3
  dim: 512
  dim: 512
}
layer {
  name: "conv1_1"
  type: "Convolution"
  bottom: "data"
  top: "conv1_1"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 64
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
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: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 64
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "relu1_2"
  type: "ReLU"
  bottom: "conv1_2"
  top: "conv1_2"
}
layer {
  name: "pool1"
  type: "Pooling"
  bottom: "conv1_2"
  top: "pool1"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "conv2_1"
  type: "Convolution"
  bottom: "pool1"
  top: "conv2_1"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 128
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
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: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 128
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "relu2_2"
  type: "ReLU"
  bottom: "conv2_2"
  top: "conv2_2"
}
layer {
  name: "pool2"
  type: "Pooling"
  bottom: "conv2_2"
  top: "pool2"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "conv3_1"
  type: "Convolution"
  bottom: "pool2"
  top: "conv3_1"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
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: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
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: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "relu3_3"
  type: "ReLU"
  bottom: "conv3_3"
  top: "conv3_3"
}
layer {
  name: "pool3"
  type: "Pooling"
  bottom: "conv3_3"
  top: "pool3"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "conv4_1"
  type: "Convolution"
  bottom: "pool3"
  top: "conv4_1"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
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: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
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: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "relu4_3"
  type: "ReLU"
  bottom: "conv4_3"
  top: "conv4_3"
}
layer {
  name: "pool4"
  type: "Pooling"
  bottom: "conv4_3"
  top: "pool4"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "conv5_1"
  type: "Convolution"
  bottom: "pool4"
  top: "conv5_1"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
    dilation: 1
  }
}
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: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
    dilation: 1
  }
}
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: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
    dilation: 1
  }
}
layer {
  name: "relu5_3"
  type: "ReLU"
  bottom: "conv5_3"
  top: "conv5_3"
}
layer {
  name: "pool5"
  type: "Pooling"
  bottom: "conv5_3"
  top: "pool5"
  pooling_param {
    pool: MAX
    kernel_size: 3
    stride: 1
    pad: 1
  }
}
layer {
  name: "fc6"
  type: "Convolution"
  bottom: "pool5"
  top: "fc6"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 1024
    pad: 6
    kernel_size: 3
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
    dilation: 6
  }
}
layer {
  name: "relu6"
  type: "ReLU"
  bottom: "fc6"
  top: "fc6"
}
layer {
  name: "fc7"
  type: "Convolution"
  bottom: "fc6"
  top: "fc7"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 1024
    kernel_size: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "relu7"
  type: "ReLU"
  bottom: "fc7"
  top: "fc7"
}
layer {
  name: "conv6_1"
  type: "Convolution"
  bottom: "fc7"
  top: "conv6_1"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 256
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "conv6_1_relu"
  type: "ReLU"
  bottom: "conv6_1"
  top: "conv6_1"
}
layer {
  name: "conv6_2"
  type: "Convolution"
  bottom: "conv6_1"
  top: "conv6_2"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "conv6_2_relu"
  type: "ReLU"
  bottom: "conv6_2"
  top: "conv6_2"
}
layer {
  name: "conv7_1"
  type: "Convolution"
  bottom: "conv6_2"
  top: "conv7_1"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 128
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "conv7_1_relu"
  type: "ReLU"
  bottom: "conv7_1"
  top: "conv7_1"
}
layer {
  name: "conv7_2"
  type: "Convolution"
  bottom: "conv7_1"
  top: "conv7_2"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
    stride: 2
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "conv7_2_relu"
  type: "ReLU"
  bottom: "conv7_2"
  top: "conv7_2"
}
layer {
  name: "conv4_3_reduce"
  type: "Convolution"
  bottom: "conv4_3"
  top: "conv4_3_reduce"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 256
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "conv4_3_reduce_relu"
  type: "ReLU"
  bottom: "conv4_3_reduce"
  top: "conv4_3_reduce"
}
layer {
  name: "fc7_reduce"
  type: "Convolution"
  bottom: "fc7"
  top: "fc7_reduce"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 256
    pad: 0
    kernel_size: 1
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fc7_reduce_relu"
  type: "ReLU"
  bottom: "fc7_reduce"
  top: "fc7_reduce"
}
layer {
  name: "fc7_us"
  type: "Interp"
  bottom: "fc7_reduce"
  top: "fc7_us"
  interp_param {
    height: 64
    width: 64
  }
}
layer {
  name: "conv7_2_us"
  type: "Interp"
  bottom: "conv7_2"
  top: "conv7_2_us"
  interp_param {
    height: 64
    width: 64
  }
}
layer {
  name: "fea_concat"
  type: "Concat"
  bottom: "conv4_3_reduce"
  bottom: "fc7_us"
  bottom: "conv7_2_us"
  top: "fea_concat"
  concat_param {
    axis: 1
  }
}
layer {
  name: "fea_concat_bn"
  type: "BatchNorm"
  bottom: "fea_concat"
  top: "fea_concat"
}
layer {
  name: "fea_concat_bn_ds_1"
  type: "Convolution"
  bottom: "fea_concat"
  top: "fea_concat_bn_ds_1"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_1_relu"
  type: "ReLU"
  bottom: "fea_concat_bn_ds_1"
  top: "fea_concat_bn_ds_1"
}
layer {
  name: "fea_concat_bn_ds_2"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_1"
  top: "fea_concat_bn_ds_2"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 512
    pad: 1
    kernel_size: 3
    stride: 2
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_2_relu"
  type: "ReLU"
  bottom: "fea_concat_bn_ds_2"
  top: "fea_concat_bn_ds_2"
}
layer {
  name: "fea_concat_bn_ds_4"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_2"
  top: "fea_concat_bn_ds_4"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
    stride: 2
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_4_relu"
  type: "ReLU"
  bottom: "fea_concat_bn_ds_4"
  top: "fea_concat_bn_ds_4"
}
layer {
  name: "fea_concat_bn_ds_8"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_4"
  top: "fea_concat_bn_ds_8"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
    stride: 2
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_8_relu"
  type: "ReLU"
  bottom: "fea_concat_bn_ds_8"
  top: "fea_concat_bn_ds_8"
}
layer {
  name: "fea_concat_bn_ds_16"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_8"
  top: "fea_concat_bn_ds_16"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
    stride: 2
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_16_relu"
  type: "ReLU"
  bottom: "fea_concat_bn_ds_16"
  top: "fea_concat_bn_ds_16"
}
layer {
  name: "fea_concat_bn_ds_32"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_16"
  top: "fea_concat_bn_ds_32"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 3
    stride: 2
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_32_relu"
  type: "ReLU"
  bottom: "fea_concat_bn_ds_32"
  top: "fea_concat_bn_ds_32"
}
layer {
  name: "fea_concat_bn_ds_64"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_32"
  top: "fea_concat_bn_ds_64"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 256
    pad: 1
    kernel_size: 4
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_64_relu"
  type: "ReLU"
  bottom: "fea_concat_bn_ds_64"
  top: "fea_concat_bn_ds_64"
}
layer {
  name: "fea_concat_bn_ds_1_mbox_loc"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_1"
  top: "fea_concat_bn_ds_1_mbox_loc"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 16
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_1_mbox_loc_perm"
  type: "Permute"
  bottom: "fea_concat_bn_ds_1_mbox_loc"
  top: "fea_concat_bn_ds_1_mbox_loc_perm"
  permute_param {
    order: 0
    order: 2
    order: 3
    order: 1
  }
}
layer {
  name: "fea_concat_bn_ds_1_mbox_loc_flat"
  type: "Flatten"
  bottom: "fea_concat_bn_ds_1_mbox_loc_perm"
  top: "fea_concat_bn_ds_1_mbox_loc_flat"
  flatten_param {
    axis: 1
  }
}
layer {
  name: "fea_concat_bn_ds_1_mbox_conf"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_1"
  top: "fea_concat_bn_ds_1_mbox_conf"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 324
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_1_mbox_conf_perm"
  type: "Permute"
  bottom: "fea_concat_bn_ds_1_mbox_conf"
  top: "fea_concat_bn_ds_1_mbox_conf_perm"
  permute_param {
    order: 0
    order: 2
    order: 3
    order: 1
  }
}
layer {
  name: "fea_concat_bn_ds_1_mbox_conf_flat"
  type: "Flatten"
  bottom: "fea_concat_bn_ds_1_mbox_conf_perm"
  top: "fea_concat_bn_ds_1_mbox_conf_flat"
  flatten_param {
    axis: 1
  }
}
layer {
  name: "fea_concat_bn_ds_1_mbox_priorbox"
  type: "PriorBox"
  bottom: "fea_concat_bn_ds_1"
  bottom: "data"
  top: "fea_concat_bn_ds_1_mbox_priorbox"
  prior_box_param {
    min_size: 35.8400001526
    max_size: 76.8000030518
    aspect_ratio: 2.0
    flip: true
    clip: false
    variance: 0.10000000149
    variance: 0.10000000149
    variance: 0.20000000298
    variance: 0.20000000298
    offset: 0.5
  }
}
layer {
  name: "fea_concat_bn_ds_2_mbox_loc"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_2"
  top: "fea_concat_bn_ds_2_mbox_loc"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 24
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_2_mbox_loc_perm"
  type: "Permute"
  bottom: "fea_concat_bn_ds_2_mbox_loc"
  top: "fea_concat_bn_ds_2_mbox_loc_perm"
  permute_param {
    order: 0
    order: 2
    order: 3
    order: 1
  }
}
layer {
  name: "fea_concat_bn_ds_2_mbox_loc_flat"
  type: "Flatten"
  bottom: "fea_concat_bn_ds_2_mbox_loc_perm"
  top: "fea_concat_bn_ds_2_mbox_loc_flat"
  flatten_param {
    axis: 1
  }
}
layer {
  name: "fea_concat_bn_ds_2_mbox_conf"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_2"
  top: "fea_concat_bn_ds_2_mbox_conf"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 486
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_2_mbox_conf_perm"
  type: "Permute"
  bottom: "fea_concat_bn_ds_2_mbox_conf"
  top: "fea_concat_bn_ds_2_mbox_conf_perm"
  permute_param {
    order: 0
    order: 2
    order: 3
    order: 1
  }
}
layer {
  name: "fea_concat_bn_ds_2_mbox_conf_flat"
  type: "Flatten"
  bottom: "fea_concat_bn_ds_2_mbox_conf_perm"
  top: "fea_concat_bn_ds_2_mbox_conf_flat"
  flatten_param {
    axis: 1
  }
}
layer {
  name: "fea_concat_bn_ds_2_mbox_priorbox"
  type: "PriorBox"
  bottom: "fea_concat_bn_ds_2"
  bottom: "data"
  top: "fea_concat_bn_ds_2_mbox_priorbox"
  prior_box_param {
    min_size: 76.8000030518
    max_size: 153.600006104
    aspect_ratio: 2.0
    aspect_ratio: 3.0
    flip: true
    clip: false
    variance: 0.10000000149
    variance: 0.10000000149
    variance: 0.20000000298
    variance: 0.20000000298
    offset: 0.5
  }
}
layer {
  name: "fea_concat_bn_ds_4_mbox_loc"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_4"
  top: "fea_concat_bn_ds_4_mbox_loc"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 24
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_4_mbox_loc_perm"
  type: "Permute"
  bottom: "fea_concat_bn_ds_4_mbox_loc"
  top: "fea_concat_bn_ds_4_mbox_loc_perm"
  permute_param {
    order: 0
    order: 2
    order: 3
    order: 1
  }
}
layer {
  name: "fea_concat_bn_ds_4_mbox_loc_flat"
  type: "Flatten"
  bottom: "fea_concat_bn_ds_4_mbox_loc_perm"
  top: "fea_concat_bn_ds_4_mbox_loc_flat"
  flatten_param {
    axis: 1
  }
}
layer {
  name: "fea_concat_bn_ds_4_mbox_conf"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_4"
  top: "fea_concat_bn_ds_4_mbox_conf"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 486
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_4_mbox_conf_perm"
  type: "Permute"
  bottom: "fea_concat_bn_ds_4_mbox_conf"
  top: "fea_concat_bn_ds_4_mbox_conf_perm"
  permute_param {
    order: 0
    order: 2
    order: 3
    order: 1
  }
}
layer {
  name: "fea_concat_bn_ds_4_mbox_conf_flat"
  type: "Flatten"
  bottom: "fea_concat_bn_ds_4_mbox_conf_perm"
  top: "fea_concat_bn_ds_4_mbox_conf_flat"
  flatten_param {
    axis: 1
  }
}
layer {
  name: "fea_concat_bn_ds_4_mbox_priorbox"
  type: "PriorBox"
  bottom: "fea_concat_bn_ds_4"
  bottom: "data"
  top: "fea_concat_bn_ds_4_mbox_priorbox"
  prior_box_param {
    min_size: 153.600006104
    max_size: 230.399993896
    aspect_ratio: 2.0
    aspect_ratio: 3.0
    flip: true
    clip: false
    variance: 0.10000000149
    variance: 0.10000000149
    variance: 0.20000000298
    variance: 0.20000000298
    offset: 0.5
  }
}
layer {
  name: "fea_concat_bn_ds_8_mbox_loc"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_8"
  top: "fea_concat_bn_ds_8_mbox_loc"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 24
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_8_mbox_loc_perm"
  type: "Permute"
  bottom: "fea_concat_bn_ds_8_mbox_loc"
  top: "fea_concat_bn_ds_8_mbox_loc_perm"
  permute_param {
    order: 0
    order: 2
    order: 3
    order: 1
  }
}
layer {
  name: "fea_concat_bn_ds_8_mbox_loc_flat"
  type: "Flatten"
  bottom: "fea_concat_bn_ds_8_mbox_loc_perm"
  top: "fea_concat_bn_ds_8_mbox_loc_flat"
  flatten_param {
    axis: 1
  }
}
layer {
  name: "fea_concat_bn_ds_8_mbox_conf"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_8"
  top: "fea_concat_bn_ds_8_mbox_conf"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 486
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_8_mbox_conf_perm"
  type: "Permute"
  bottom: "fea_concat_bn_ds_8_mbox_conf"
  top: "fea_concat_bn_ds_8_mbox_conf_perm"
  permute_param {
    order: 0
    order: 2
    order: 3
    order: 1
  }
}
layer {
  name: "fea_concat_bn_ds_8_mbox_conf_flat"
  type: "Flatten"
  bottom: "fea_concat_bn_ds_8_mbox_conf_perm"
  top: "fea_concat_bn_ds_8_mbox_conf_flat"
  flatten_param {
    axis: 1
  }
}
layer {
  name: "fea_concat_bn_ds_8_mbox_priorbox"
  type: "PriorBox"
  bottom: "fea_concat_bn_ds_8"
  bottom: "data"
  top: "fea_concat_bn_ds_8_mbox_priorbox"
  prior_box_param {
    min_size: 230.399993896
    max_size: 307.200012207
    aspect_ratio: 2.0
    aspect_ratio: 3.0
    flip: true
    clip: false
    variance: 0.10000000149
    variance: 0.10000000149
    variance: 0.20000000298
    variance: 0.20000000298
    offset: 0.5
  }
}
layer {
  name: "fea_concat_bn_ds_16_mbox_loc"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_16"
  top: "fea_concat_bn_ds_16_mbox_loc"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 24
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_16_mbox_loc_perm"
  type: "Permute"
  bottom: "fea_concat_bn_ds_16_mbox_loc"
  top: "fea_concat_bn_ds_16_mbox_loc_perm"
  permute_param {
    order: 0
    order: 2
    order: 3
    order: 1
  }
}
layer {
  name: "fea_concat_bn_ds_16_mbox_loc_flat"
  type: "Flatten"
  bottom: "fea_concat_bn_ds_16_mbox_loc_perm"
  top: "fea_concat_bn_ds_16_mbox_loc_flat"
  flatten_param {
    axis: 1
  }
}
layer {
  name: "fea_concat_bn_ds_16_mbox_conf"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_16"
  top: "fea_concat_bn_ds_16_mbox_conf"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 486
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_16_mbox_conf_perm"
  type: "Permute"
  bottom: "fea_concat_bn_ds_16_mbox_conf"
  top: "fea_concat_bn_ds_16_mbox_conf_perm"
  permute_param {
    order: 0
    order: 2
    order: 3
    order: 1
  }
}
layer {
  name: "fea_concat_bn_ds_16_mbox_conf_flat"
  type: "Flatten"
  bottom: "fea_concat_bn_ds_16_mbox_conf_perm"
  top: "fea_concat_bn_ds_16_mbox_conf_flat"
  flatten_param {
    axis: 1
  }
}
layer {
  name: "fea_concat_bn_ds_16_mbox_priorbox"
  type: "PriorBox"
  bottom: "fea_concat_bn_ds_16"
  bottom: "data"
  top: "fea_concat_bn_ds_16_mbox_priorbox"
  prior_box_param {
    min_size: 307.200012207
    max_size: 384.0
    aspect_ratio: 2.0
    aspect_ratio: 3.0
    flip: true
    clip: false
    variance: 0.10000000149
    variance: 0.10000000149
    variance: 0.20000000298
    variance: 0.20000000298
    offset: 0.5
  }
}
layer {
  name: "fea_concat_bn_ds_32_mbox_loc"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_32"
  top: "fea_concat_bn_ds_32_mbox_loc"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 16
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_32_mbox_loc_perm"
  type: "Permute"
  bottom: "fea_concat_bn_ds_32_mbox_loc"
  top: "fea_concat_bn_ds_32_mbox_loc_perm"
  permute_param {
    order: 0
    order: 2
    order: 3
    order: 1
  }
}
layer {
  name: "fea_concat_bn_ds_32_mbox_loc_flat"
  type: "Flatten"
  bottom: "fea_concat_bn_ds_32_mbox_loc_perm"
  top: "fea_concat_bn_ds_32_mbox_loc_flat"
  flatten_param {
    axis: 1
  }
}
layer {
  name: "fea_concat_bn_ds_32_mbox_conf"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_32"
  top: "fea_concat_bn_ds_32_mbox_conf"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 324
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_32_mbox_conf_perm"
  type: "Permute"
  bottom: "fea_concat_bn_ds_32_mbox_conf"
  top: "fea_concat_bn_ds_32_mbox_conf_perm"
  permute_param {
    order: 0
    order: 2
    order: 3
    order: 1
  }
}
layer {
  name: "fea_concat_bn_ds_32_mbox_conf_flat"
  type: "Flatten"
  bottom: "fea_concat_bn_ds_32_mbox_conf_perm"
  top: "fea_concat_bn_ds_32_mbox_conf_flat"
  flatten_param {
    axis: 1
  }
}
layer {
  name: "fea_concat_bn_ds_32_mbox_priorbox"
  type: "PriorBox"
  bottom: "fea_concat_bn_ds_32"
  bottom: "data"
  top: "fea_concat_bn_ds_32_mbox_priorbox"
  prior_box_param {
    min_size: 384.0
    max_size: 460.799987793
    aspect_ratio: 2.0
    flip: true
    clip: false
    variance: 0.10000000149
    variance: 0.10000000149
    variance: 0.20000000298
    variance: 0.20000000298
    offset: 0.5
  }
}
layer {
  name: "fea_concat_bn_ds_64_mbox_loc"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_64"
  top: "fea_concat_bn_ds_64_mbox_loc"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 16
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_64_mbox_loc_perm"
  type: "Permute"
  bottom: "fea_concat_bn_ds_64_mbox_loc"
  top: "fea_concat_bn_ds_64_mbox_loc_perm"
  permute_param {
    order: 0
    order: 2
    order: 3
    order: 1
  }
}
layer {
  name: "fea_concat_bn_ds_64_mbox_loc_flat"
  type: "Flatten"
  bottom: "fea_concat_bn_ds_64_mbox_loc_perm"
  top: "fea_concat_bn_ds_64_mbox_loc_flat"
  flatten_param {
    axis: 1
  }
}
layer {
  name: "fea_concat_bn_ds_64_mbox_conf"
  type: "Convolution"
  bottom: "fea_concat_bn_ds_64"
  top: "fea_concat_bn_ds_64_mbox_conf"
  param {
    lr_mult: 1.0
    decay_mult: 1.0
  }
  param {
    lr_mult: 2.0
    decay_mult: 0.0
  }
  convolution_param {
    num_output: 324
    pad: 1
    kernel_size: 3
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0.0
    }
  }
}
layer {
  name: "fea_concat_bn_ds_64_mbox_conf_perm"
  type: "Permute"
  bottom: "fea_concat_bn_ds_64_mbox_conf"
  top: "fea_concat_bn_ds_64_mbox_conf_perm"
  permute_param {
    order: 0
    order: 2
    order: 3
    order: 1
  }
}
layer {
  name: "fea_concat_bn_ds_64_mbox_conf_flat"
  type: "Flatten"
  bottom: "fea_concat_bn_ds_64_mbox_conf_perm"
  top: "fea_concat_bn_ds_64_mbox_conf_flat"
  flatten_param {
    axis: 1
  }
}
layer {
  name: "fea_concat_bn_ds_64_mbox_priorbox"
  type: "PriorBox"
  bottom: "fea_concat_bn_ds_64"
  bottom: "data"
  top: "fea_concat_bn_ds_64_mbox_priorbox"
  prior_box_param {
    min_size: 460.799987793
    max_size: 537.599975586
    aspect_ratio: 2.0
    flip: true
    clip: false
    variance: 0.10000000149
    variance: 0.10000000149
    variance: 0.20000000298
    variance: 0.20000000298
    offset: 0.5
  }
}
layer {
  name: "mbox_loc"
  type: "Concat"
  bottom: "fea_concat_bn_ds_1_mbox_loc_flat"
  bottom: "fea_concat_bn_ds_2_mbox_loc_flat"
  bottom: "fea_concat_bn_ds_4_mbox_loc_flat"
  bottom: "fea_concat_bn_ds_8_mbox_loc_flat"
  bottom: "fea_concat_bn_ds_16_mbox_loc_flat"
  bottom: "fea_concat_bn_ds_32_mbox_loc_flat"
  bottom: "fea_concat_bn_ds_64_mbox_loc_flat"
  top: "mbox_loc"
  concat_param {
    axis: 1
  }
}
layer {
  name: "mbox_conf"
  type: "Concat"
  bottom: "fea_concat_bn_ds_1_mbox_conf_flat"
  bottom: "fea_concat_bn_ds_2_mbox_conf_flat"
  bottom: "fea_concat_bn_ds_4_mbox_conf_flat"
  bottom: "fea_concat_bn_ds_8_mbox_conf_flat"
  bottom: "fea_concat_bn_ds_16_mbox_conf_flat"
  bottom: "fea_concat_bn_ds_32_mbox_conf_flat"
  bottom: "fea_concat_bn_ds_64_mbox_conf_flat"
  top: "mbox_conf"
  concat_param {
    axis: 1
  }
}
layer {
  name: "mbox_priorbox"
  type: "Concat"
  bottom: "fea_concat_bn_ds_1_mbox_priorbox"
  bottom: "fea_concat_bn_ds_2_mbox_priorbox"
  bottom: "fea_concat_bn_ds_4_mbox_priorbox"
  bottom: "fea_concat_bn_ds_8_mbox_priorbox"
  bottom: "fea_concat_bn_ds_16_mbox_priorbox"
  bottom: "fea_concat_bn_ds_32_mbox_priorbox"
  bottom: "fea_concat_bn_ds_64_mbox_priorbox"
  top: "mbox_priorbox"
  concat_param {
    axis: 2
  }
}
layer {
  name: "mbox_conf_reshape"
  type: "Reshape"
  bottom: "mbox_conf"
  top: "mbox_conf_reshape"
  reshape_param {
    shape {
      dim: 0
      dim: -1
      dim: 81
    }
  }
}
layer {
  name: "mbox_conf_softmax"
  type: "Softmax"
  bottom: "mbox_conf_reshape"
  top: "mbox_conf_softmax"
  softmax_param {
    axis: 2
  }
}
layer {
  name: "mbox_conf_flatten"
  type: "Flatten"
  bottom: "mbox_conf_softmax"
  top: "mbox_conf_flatten"
  flatten_param {
    axis: 1
  }
}
layer {
  name: "detection_out"
  type: "DetectionOutput"
  bottom: "mbox_loc"
  bottom: "mbox_conf_flatten"
  bottom: "mbox_priorbox"
  top: "detection_out"
  include {
    phase: TEST
  }
  detection_output_param {
    num_classes: 81
    share_location: true
    background_label_id: 0
    nms_param {
      nms_threshold: 0.449999988079
      top_k: 400
    }
    save_output_param {
      output_directory: "jobs/VGGNet/SSD_FPN_NO33_WOP512x512/1025/predict_ss"
      output_name_prefix: "detections_minival_fssd512_"
      output_format: "COCO"
      label_map_file: "data/coco/labelmap_coco.prototxt"
      name_size_file: "data/coco/minival2014_name_size.txt"
      num_test_image: 5000
    }
    code_type: CENTER_SIZE
    keep_top_k: 200
    confidence_threshold: 0.00999999977648
  }
}
@dkurt dkurt mentioned this issue Mar 12, 2018
Merged
@dkurt dkurt self-assigned this Mar 19, 2018
@dkurt dkurt mentioned this issue Mar 21, 2018
Merged
@dkurt
Copy link
Member

dkurt commented Mar 21, 2018

@fabito, Many thanks! Looks like Interp layer does not match with any OpenCV's resize interpolation. I'm going to suggest you try custom layers introduced at #11129. You may find Interp layer implementation in the tutorial there.

@dkurt
Copy link
Member

dkurt commented May 4, 2018

Please see a tutorial https://docs.opencv.org/master/dc/db1/tutorial_dnn_custom_layers.html.

@dkurt dkurt closed this as completed May 4, 2018
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@dkurt dkurt

Labels
category: dnn feature
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@fabito @dkurt