精度始终为 1 Caffe 回归

Question

我的数据集包含 400 张 32x32x3 的图像，标签包含浮点数 (-1,1)。示例：

faceCroppedImages/img1.jpg 0
faceCroppedImages/img2.jpg 0.0128
faceCroppedImages/img3.jpg 0.0128
faceCroppedImages/img4.jpg 0.0128
faceCroppedImages/img22.jpg 0.0128
faceCroppedImages/img23.jpg 0.0085
faceCroppedImages/img24.jpg 0.0077
faceCroppedImages/img25.jpg 0.0077
faceCroppedImages/img293.jpg -0.023
faceCroppedImages/img294.jpg -0.023
faceCroppedImages/img295.jpg -0.0204
faceCroppedImages/img296.jpg -0.0179
faceCroppedImages/img297.jpg -0.017
faceCroppedImages/img298.jpg -0.0128

我的'solver.prototxt'是：

net: "train_test_hdf5.prototxt"
test_iter: 100
test_interval: 500
base_lr: 0.003
momentum: 0.9
weight_decay: 0.0005
lr_policy: "inv"
gamma: 0.0001
power: 0.75
display: 100
max_iter: 10000
snapshot: 5000
snapshot_prefix: "lenet_hdf5"
solver_mode: CPU

和'train_test_hdf5.prototxt'是：

name: "MSE regression"
layer{
  name: "data"
  type: "HDF5Data"
  top: "data"
  top: "label"
  hdf5_data_param {
    source: "train_hdf5file.txt"
    batch_size: 64
    shuffle: true
  }
  include: { phase: TRAIN }
}

layer{
  name: "data"
  type: "HDF5Data"
  top: "data"
  top: "label"
  hdf5_data_param {
    source: "test_hdf5file.txt"
    batch_size: 128
  }
  include: { phase: TEST }
}
layer {
  name: "conv1"
  type: "Convolution"
  bottom: "data"
  top: "conv1"
  param { lr_mult: 1 }
  param { lr_mult: 2 }
  convolution_param {
    num_output: 20
    kernel_size: 5
    stride: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "relu1"
  type: "ReLU"
  bottom: "conv1"
  top: "conv1"
}
layer {
  name: "pool1"
  type: "Pooling"
  bottom: "conv1"
  top: "pool1"
  pooling_param {
    pool: MAX
    kernel_size: 2
    stride: 2
  }
}
layer {
  name: "dropout1"
  type: "Dropout"
  bottom: "pool1"
  top: "pool1"
  dropout_param {
    dropout_ratio: 0.1
  }
}

layer{
  name: "fc1"
  type: "InnerProduct"
  bottom: "pool1"
  top: "fc1"
  param { lr_mult: 1 decay_mult: 1 }
  param { lr_mult: 2 decay_mult: 0 }
  inner_product_param {
    num_output: 500
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0
    }
  }
}
layer {
  name: "dropout2"
  type: "Dropout"
  bottom: "fc1"
  top: "fc1"
  dropout_param {
    dropout_ratio: 0.5
  }
}
layer{
  name: "fc2"
  type: "InnerProduct"
  bottom: "fc1"
  top: "fc2"
  param { lr_mult: 1 decay_mult: 1 }
  param { lr_mult: 2 decay_mult: 0 }
  inner_product_param {
    num_output: 1
    weight_filler {
      type: "xavier"
    }
    bias_filler {
      type: "constant"
      value: 0
    }
 }
}
layer {
  name: "accuracy1"
  type: "Accuracy"
  bottom: "fc2"
  bottom: "label"
  top: "accuracy1"
  include {
    phase: TEST
  }
}
layer{
  name: "loss"
  type: "EuclideanLoss"
  bottom: "fc2"
  bottom: "label"
  top: "loss"
}

然而，当我测试数据时，准确度始终为 1:

我尝试使用整数标签将当前标签乘以 1000，但出现 nan 错误：

你能告诉我哪里做错了吗？我是咖啡和神经网络的初学者。任何建议都是有价值的。 TIA.

Answer 1

使用 "Accuracy" 层进行回归任务没有意义：该层测量 classification 输出的准确性。
例如，如果您尝试预测 L 标签之一，fc2 层的 num_output 将是 L - 也就是说，预测每个 class 的概率.然后 "Accuracy" 层检查对应于预期输出 l 的第 l 个条目的概率是否最大。
当 fc2 输出只有一维时，你怎么能计算出这样的精度？

在您的情况下，您只能检查欧几里德损失并看到它在测试和训练中都在下降。

精度始终为 1 Caffe 回归

Accuracy always 1 Caffe Regression

regression

neural-network

deep-learning

caffe

conv-neural-network