检查失败：axis_index <num_axes（）（1 vs.0）轴1超出形状为（0）的0-D Blob的范围

Question

当我从GitHub Face_Alignment_Two_Stage_Re-initialization运行代码时，出现以下错误：

I0911 08:08:29.921020 12228 layer_factory.hpp:88] Creating layer input
I0911 08:08:29.922020 12228 net.cpp:100] Creating Layer input
I0911 08:08:29.922020 12228 net.cpp:418] input -> data
I0911 08:08:29.923020 12228 net.cpp:150] Setting up input
I0911 08:08:29.924021 12228 net.cpp:157] Top shape: 1 3 448 448 (602112)
I0911 08:08:29.927022 12228 net.cpp:165] Memory required for data: 2408448
I0911 08:08:29.927022 12228 layer_factory.hpp:88] Creating layer data_input_0_split
I0911 08:08:29.928025 12228 net.cpp:100] Creating Layer data_input_0_split
I0911 08:08:29.928025 12228 net.cpp:444] data_input_0_split <- data
I0911 08:08:29.929024 12228 net.cpp:418] data_input_0_split -> data_input_0_split_0
I0911 08:08:29.930024 12228 net.cpp:418] data_input_0_split -> data_input_0_split_1
I0911 08:08:29.930024 12228 net.cpp:150] Setting up data_input_0_split
I0911 08:08:29.930024 12228 net.cpp:157] Top shape: 1 3 448 448 (602112)
I0911 08:08:29.932025 12228 net.cpp:157] Top shape: 1 3 448 448 (602112)
I0911 08:08:29.932025 12228 net.cpp:165] Memory required for data: 7225344
I0911 08:08:29.932025 12228 layer_factory.hpp:88] Creating layer downsample_data
I0911 08:08:29.932025 12228 net.cpp:100] Creating Layer downsample_data
I0911 08:08:29.932025 12228 net.cpp:444] downsample_data <- data_input_0_split_0
I0911 08:08:29.932025 12228 net.cpp:418] downsample_data -> downsample_data
I0911 08:08:29.932025 12228 net.cpp:150] Setting up downsample_data
I0911 08:08:29.932025 12228 net.cpp:157] Top shape: (0)
I0911 08:08:29.932025 12228 net.cpp:165] Memory required for data: 7225344
I0911 08:08:29.933027 12228 layer_factory.hpp:88] Creating layer net1_conv1
I0911 08:08:29.933027 12228 net.cpp:100] Creating Layer net1_conv1
I0911 08:08:29.933027 12228 net.cpp:444] net1_conv1 <- downsample_data
I0911 08:08:29.933027 12228 net.cpp:418] net1_conv1 -> net1_conv1
F0911 08:08:29.933027 12228 blob.hpp:122] Check failed: axis_index < num_axes() (1 vs. 0) axis 1 out of range for 0-D Blob with shape (0)
*** Check failure stack trace: ***

似乎是由于SubsamplePooling层而发生了错误，为什么顶部形状为0？我认为SubsamplePooling层的顶部形状应为1 3 60 60。 部署的部分如下。

name: "facial_point_net"
input: "data"
input_dim: 1
input_dim: 3
input_dim: 448
input_dim: 448

layer {
  name: "downsample_data"
  type: "SubsamplePooling"
  bottom: "data"
  top: "downsample_data"
  subsample_pooling_param {
    output_H: 60
    output_W: 60
  }
}
##########################################
layer {
  name: "net1_conv1"
  type: "Convolution"
  bottom: "downsample_data"
  top: "net1_conv1"
  param {
    lr_mult: 1
    decay_mult: 1
  }
  param {
    lr_mult: 2
    decay_mult: 0
  }
  convolution_param {
    num_output: 20
    pad: 0
    kernel_size: 5
    stride: 1
    weight_filler {
      type: "gaussian"
      std: 0.01
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "net1_PReLU1"
  type: "PReLU"
  bottom: "net1_conv1"
  top: "net1_conv1"
}

layer {
  name: "net1_pool1"
  type: "Pooling"
  bottom: "net1_conv1"
  top: "net1_pool1"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}

有谁知道是什么问题？我需要帮助。