CNN 构建例子by Tensorflow 2.0

CNN 结构

构建CNN的结构参考了下面的图：

图片来自于论文：
T. Klamt and S. Behnke, “Towards Learning Abstract Representations for Locomotion Planning in High-dimensional State Spaces,” 2019 International Conference on Robotics and Automation (ICRA), Montreal, QC, Canada, 2019, pp. 922-928.
doi: 10.1109/ICRA.2019.8794144

Tensorflow 代码：

from __future__ import absolute_import, division, print_function, unicode_literals
import tensorflow as tf
from tensorflow.keras import datasets, layers, models
import matplotlib.pyplot as plt


# import the figure


model = models.Sequential()
# model.add(layers.Conv2D(32, (3, 3), activation='relu', input_shape=(32, 32, 3) ))
# model.add(layers.Conv2D(input_shape=(x_train.shape[1], x_train.shape[2], x_train.shape[3]),
#                         filters=32, kernel_size=(3,3), strides=(1,1), padding='valid',
#                        activation='relu'))

# (1)CONV layer, f*f =3*3, padding = 1, stride = 1
model.add(layers.Conv2D(input_shape=(72, 72, 1),
                       filters=3, kernel_size=(3,3),strides=1,padding='same',
                       activation='relu'))
# (2)CONV layer, f*f = 7*7, padding = 3=(f-1)/2, stride = 1
model.add(layers.Conv2D(filters=5, kernel_size=(7,7),strides=(1,1),padding='same',
                       activation='relu'))

# (3)CONV layer, f*f = 14*14, padding =0, stride = 1
model.add(layers.Conv2D(filters=28, kernel_size=(14,14),strides=(1,1),padding='valid',
                       activation='relu'))
# (4)CONV layer, f*f = 4*4, padding = 0, stride =1, max pooling, f=2,s=2
model.add(layers.Conv2D(filters=31, kernel_size=(4,4),strides=(1,1),padding='valid',
                       activation='relu'))
# model.add(layers.MaxPool2D(pool_size(2, 2)))
model.add(layers.MaxPool2D((2, 2)))

# (5)CONV layer, f*f = 3*3, padding = 0, stride=1, maxpooling
model.add(layers.Conv2D(filters=34, kernel_size=(3,3),strides=(1,1),padding='valid',
                        activation='relu'))
model.add(layers.MaxPool2D((2, 2)))

# # (6) CONV layer, f*f = 3*3, p=0, s=1
model.add(layers.Conv2D(filters=36, kernel_size=(3,3),strides=(1,1),padding='valid',
                        activation='relu'))
# # (7)CONV layer, f*f = 3*3, p=0,s=1
model.add(layers.Conv2D(filters=38, kernel_size=(3,3),strides=(1,1),padding='valid',
                        activation='relu'))
# # (8)CONV layer, f*f = 3*3, p=0, s=1
model.add(layers.Conv2D(filters=40, kernel_size=(3,3),strides=(1,1),padding='valid',
                        activation='relu'))
#
# # (n)Dense layer
# #To complete our model, you will feed the last output tensor
# # from the convolutional base (of shape (3, 3, 64)) into one or more Dense layers to perform classification.
# #Dense layers take vectors as input (which are 1D), while the current output is a 3D tensor.
# #First, you will flatten (or unroll) the 3D output to 1D, then add one or more Dense layers on top.
# #CIFAR has 10 output classes, so you use a final Dense layer with 10 outputs and a softmax activation.
model.add(layers.Flatten())
model.add(layers.Dense(500, activation='relu')) # this should equal the filters number for last layer.
model.add(layers.Dense(20, activation='softmax')) # 20: the number of outputs
#
model.summary()

关于格式，可以参考tensorflow_core/python/keras/layers/convolutional.py 文件， 里面关于Conv2 的说明如下：

@keras_export('keras.layers.Conv2D', 'keras.layers.Convolution2D')
class Conv2D(Conv):
  """2D convolution layer (e.g. spatial convolution over images).

  This layer creates a convolution kernel that is convolved
  with the layer input to produce a tensor of
  outputs. If `use_bias` is True,
  a bias vector is created and added to the outputs. Finally, if
  `activation` is not `None`, it is applied to the outputs as well.

  When using this layer as the first layer in a model,
  provide the keyword argument `input_shape`
  (tuple of integers, does not include the sample axis),
  e.g. `input_shape=(128, 128, 3)` for 128x128 RGB pictures
  in `data_format="channels_last"`.

  Arguments:
    filters: Integer, the dimensionality of the output space
      (i.e. the number of output filters in the convolution).
    kernel_size: An integer or tuple/list of 2 integers, specifying the
      height and width of the 2D convolution window.
      Can be a single integer to specify the same value for
      all spatial dimensions.
    strides: An integer or tuple/list of 2 integers,
      specifying the strides of the convolution along the height and width.
      Can be a single integer to specify the same value for
      all spatial dimensions.
      Specifying any stride value != 1 is incompatible with specifying
      any `dilation_rate` value != 1.
    padding: one of `"valid"` or `"same"` (case-insensitive).
    data_format: A string,
      one of `channels_last` (default) or `channels_first`.
      The ordering of the dimensions in the inputs.
      `channels_last` corresponds to inputs with shape
      `(batch, height, width, channels)` while `channels_first`
      corresponds to inputs with shape
      `(batch, channels, height, width)`.
      It defaults to the `image_data_format` value found in your
      Keras config file at `~/.keras/keras.json`.
      If you never set it, then it will be "channels_last".
    dilation_rate: an integer or tuple/list of 2 integers, specifying
      the dilation rate to use for dilated convolution.
      Can be a single integer to specify the same value for
      all spatial dimensions.
      Currently, specifying any `dilation_rate` value != 1 is
      incompatible with specifying any stride value != 1.
    activation: Activation function to use.
      If you don't specify anything, no activation is applied
      (ie. "linear" activation: `a(x) = x`).
    use_bias: Boolean, whether the layer uses a bias vector.
    kernel_initializer: Initializer for the `kernel` weights matrix.
    bias_initializer: Initializer for the bias vector.
    kernel_regularizer: Regularizer function applied to
      the `kernel` weights matrix.
    bias_regularizer: Regularizer function applied to the bias vector.
    activity_regularizer: Regularizer function applied to
      the output of the layer (its "activation")..
    kernel_constraint: Constraint function applied to the kernel matrix.
    bias_constraint: Constraint function applied to the bias vector.

  Input shape:
    4D tensor with shape:
    `(samples, channels, rows, cols)` if data_format='channels_first'
    or 4D tensor with shape:
    `(samples, rows, cols, channels)` if data_format='channels_last'.

  Output shape:
    4D tensor with shape:
    `(samples, filters, new_rows, new_cols)` if data_format='channels_first'
    or 4D tensor with shape:
    `(samples, new_rows, new_cols, filters)` if data_format='channels_last'.
    `rows` and `cols` values might have changed due to padding.
  """

  def __init__(self,
               filters,
               kernel_size,
               strides=(1, 1),
               padding='valid',
               data_format=None,
               dilation_rate=(1, 1),
               activation=None,
               use_bias=True,
               kernel_initializer='glorot_uniform',
               bias_initializer='zeros',
               kernel_regularizer=None,
               bias_regularizer=None,
               activity_regularizer=None,
               kernel_constraint=None,
               bias_constraint=None,
               **kwargs):
    super(Conv2D, self).__init__(
        rank=2,
        filters=filters,
        kernel_size=kernel_size,
        strides=strides,
        padding=padding,
        data_format=data_format,
        dilation_rate=dilation_rate,
        activation=activations.get(activation),
        use_bias=use_bias,
        kernel_initializer=initializers.get(kernel_initializer),
        bias_initializer=initializers.get(bias_initializer),
        kernel_regularizer=regularizers.get(kernel_regularizer),
        bias_regularizer=regularizers.get(bias_regularizer),
        activity_regularizer=regularizers.get(activity_regularizer),
        kernel_constraint=constraints.get(kernel_constraint),
        bias_constraint=constraints.get(bias_constraint),
        **kwargs)

运行结果

Model: "sequential"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
conv2d (Conv2D)              (None, 72, 72, 3)         30        
_________________________________________________________________
conv2d_1 (Conv2D)            (None, 72, 72, 5)         740       
_________________________________________________________________
conv2d_2 (Conv2D)            (None, 59, 59, 28)        27468     
_________________________________________________________________
conv2d_3 (Conv2D)            (None, 56, 56, 31)        13919     
_________________________________________________________________
max_pooling2d (MaxPooling2D) (None, 28, 28, 31)        0         
_________________________________________________________________
conv2d_4 (Conv2D)            (None, 26, 26, 34)        9520      
_________________________________________________________________
max_pooling2d_1 (MaxPooling2 (None, 13, 13, 34)        0         
_________________________________________________________________
conv2d_5 (Conv2D)            (None, 11, 11, 36)        11052     
_________________________________________________________________
conv2d_6 (Conv2D)            (None, 9, 9, 38)          12350     
_________________________________________________________________
conv2d_7 (Conv2D)            (None, 7, 7, 40)          13720     
_________________________________________________________________
flatten (Flatten)            (None, 1960)              0         
_________________________________________________________________
dense (Dense)                (None, 500)               980500    
_________________________________________________________________
dense_1 (Dense)              (None, 20)                10020     
=================================================================
Total params: 1,079,319
Trainable params: 1,079,319
Non-trainable params: 0
_________________________________________________________________

Process finished with exit code 0