(1)tf.nn.max_pool()函數
解釋:
tf.nn.max_pool(value, ksize, strides, padding, data_format='NHWC', name=None) 需要設置的參數主要有四個: 第一個參數value:需要池化的輸入,一般池化層接在卷積層后面,所以輸入通常是feature map,依然是[batch, height, width, channels]這樣的shape 第二個參數ksize:池化窗口的大小,取一個四維向量,一般是[1, height, width, 1],因為我們不想在batch和channels上做池化,所以這兩個維度設為了1 第三個參數strides:和卷積類似,窗口在每一個維度上滑動的步長,一般也是[1, stride,stride, 1] 第四個參數padding:和卷積類似,可以取'VALID' 或者'SAME' 返回一個Tensor,類型不變,shape仍然是[batch, height, width, channels]這種形式
示例:
程序: import tensorflow as tf a=tf.constant([ [[1.0,2.0,3.0,4.0], [5.0,6.0,7.0,8.0], [8.0,7.0,6.0,5.0], [4.0,3.0,2.0,1.0]], [[4.0,3.0,2.0,1.0], [8.0,7.0,6.0,5.0], [1.0,2.0,3.0,4.0], [5.0,6.0,7.0,8.0]] ]) a=tf.reshape(a,[1,4,4,2]) pooling=tf.nn.max_pool(a,[1,2,2,1],[1,1,1,1],padding='VALID') with tf.Session() as sess: print("image:") image=sess.run(a) print (image) print("reslut:") result=sess.run(pooling) print (result) 運行結果: image: [[[[ 1. 2.] [ 3. 4.] [ 5. 6.] [ 7. 8.]] [[ 8. 7.] [ 6. 5.] [ 4. 3.] [ 2. 1.]] [[ 4. 3.] [ 2. 1.] [ 8. 7.] [ 6. 5.]] [[ 1. 2.] [ 3. 4.] [ 5. 6.] [ 7. 8.]]]] reslut: [[[[ 8. 7.] [ 6. 6.] [ 7. 8.]] [[ 8. 7.] [ 8. 7.] [ 8. 7.]] [[ 4. 4.] [ 8. 7.] [ 8. 8.]]]]
(2)tf.nn.dropout函數
解釋:
tf.nn.dropout(x, keep_prob, noise_shape=None, seed=None, name=None) 此函數是為了防止在訓練中過擬合的操作,將訓練輸出按一定規則進行變換 參數: x:輸入 keep_prob:保留比例。 取值 (0,1] 。每一個參數都將按這個比例隨機變更 noise_shape:干擾形狀。 此字段默認是None,表示第一個元素的操作都是獨立,但是也不一定。比例:數據的形狀是shape(x)=[k, l, m, n],而noise_shape=[k, 1, 1, n],則第1和4列是獨立保留或刪除,第2和3列是要么全部保留,要么全部刪除。 seed:整形變量,隨機數種子。 name:名字,沒啥用 返回:Tnesor
(3)tf.nn.local_response_normalization函數
公式說明
local response normalization最早是由Krizhevsky和Hinton在關於ImageNet的論文里面使用的一種數據標准化方法,即使現在,也依然會有不少CNN網絡會使用到這種正則手段,現在記錄一下lrn方法的計算流程以及tensorflow的實現,方便后面查閱
以上是這種歸一手段的公式,其中a的上標指該層的第幾個feature map,a的下標x,y表示feature map的像素位置,N指feature map的總數量,公式里的其它參數都是超參,需要自己指定的。
這種方法是受到神經科學的啟發,激活的神經元會抑制其鄰近神經元的活動(側抑制現象),至於為什么使用這種正則手段,以及它為什么有效,查閱了很多文獻似乎也沒有詳細的解釋,可能是由於后來提出的batch normalization手段太過火熱,漸漸的就把local response normalization掩蓋了吧
解釋:
tf.nn.local_response_normalization(input, depth_radius=5, bias=1, alpha=1, beta=0.5, name=None) 除去name參數用以指定該操作的name,與方法有關的一共五個參數: 第一個參數input:這個輸入就是feature map了,既然是feature map,那么它就具有[batch, height, width, channels]這樣的shape 第二個參數depth_radius:這個值需要自己指定,就是上述公式中的n/2 第三個參數bias:上述公式中的k 第四個參數alpha:上述公式中的α 第五個參數beta:上述公式中的β
程序:
import tensorflow as tf a = tf.constant([ [[1.0, 2.0, 3.0, 4.0], [5.0, 6.0, 7.0, 8.0], [8.0, 7.0, 6.0, 5.0], [4.0, 3.0, 2.0, 1.0]], [[4.0, 3.0, 2.0, 1.0], [8.0, 7.0, 6.0, 5.0], [1.0, 2.0, 3.0, 4.0], [5.0, 6.0, 7.0, 8.0]] ]) #reshape a,get the feature map [batch:1 height:2 width:2 channels:8] a = tf.reshape(a, [1, 2, 2, 8]) normal_a=tf.nn.local_response_normalization(a,2,0,1,1) with tf.Session() as sess: print("feature map:") image = sess.run(a) print (image) print("normalized feature map:") normal = sess.run(normal_a) print (normal)
輸出結果:
feature map: [[[[ 1. 2. 3. 4. 5. 6. 7. 8.] [ 8. 7. 6. 5. 4. 3. 2. 1.]] [[ 4. 3. 2. 1. 8. 7. 6. 5.] [ 1. 2. 3. 4. 5. 6. 7. 8.]]]] normalized feature map: [[[[ 0.07142857 0.06666667 0.05454545 0.04444445 0.03703704 0.03157895 0.04022989 0.05369128] [ 0.05369128 0.04022989 0.03157895 0.03703704 0.04444445 0.05454545 0.06666667 0.07142857]] [[ 0.13793103 0.10000001 0.0212766 0.00787402 0.05194805 0.04 0.03448276 0.04545454] [ 0.07142857 0.06666667 0.05454545 0.04444445 0.03703704 0.03157895 0.04022989 0.05369128]]]]
(4)tf.get_variable函數
函數定義:
get_variable( name, shape=None, dtype=None, initializer=None, regularizer=None, trainable=True, collections=None, caching_device=None, partitioner=None, validate_shape=True, use_resource=None, custom_getter=None )
其中參數分別為:
參數: name:新變量或現有變量的名稱。 shape:新變量或現有變量的形狀。 dtype:新變量或現有變量的類型(默認為 DT_FLOAT)。 initializer:創建變量的初始化器。 regularizer:一個函數(張量 - >張量或無);將其應用於新創建的變量的結果將被添加到集合 tf.GraphKeys.REGULARIZATION_LOSSES 中,並可用於正則化。 trainable:如果為 True,還將變量添加到圖形集合:GraphKeys.TRAINABLE_VARIABLES。 collections:要將變量添加到其中的圖形集合鍵的列表。默認為 [GraphKeys.LOCAL_VARIABLES]。 caching_device:可選的設備字符串或函數,描述變量應該被緩存以讀取的位置。默認為變量的設備,如果不是 None,則在其他設備上進行緩存。典型的用法的在使用該變量的操作所在的設備上進行緩存,通過 Switch 和其他條件語句來復制重復數據刪除。 partitioner:(可選)可調用性,它接受要創建的變量的完全定義的 TensorShape 和 dtype,並且返回每個坐標軸的分區列表(當前只能對一個坐標軸進行分區)。 validate_shape:如果為假,則允許使用未知形狀的值初始化變量。如果為真,則默認情況下,initial_value 的形狀必須是已知的。 use_resource:如果為假,則創建一個常規變量。如果為真,則創建一個實驗性的 ResourceVariable,而不是具有明確定義的語義。默認為假(稍后將更改為真)。 custom_getter:可調用的,將第一個參數作為真正的 getter,並允許覆蓋內部的 get_variable 方法。custom_getter 的簽名應該符合這種方法,但最經得起未來考驗的版本將允許更改:def custom_getter(getter, *args, **kwargs)。還允許直接訪問所有 get_variable 參數:def custom_getter(getter, name, *args, **kwargs)。創建具有修改的名稱的變量的簡單標識自定義 getter 是:python def custom_getter(getter, name, *args, **kwargs): return getter(name + '_suffix', *args, **kwargs)
使用例子:
w = tf.get_variable("w", shape = [inputD, outputD], dtype = "float") b = tf.get_variable("b", [outputD], dtype = "float")
(5)tf.variable_scope函數
函數原理
用於定義創建變量(層)的操作的上下文管理器。
此上下文管理器驗證(可選)values是否來自同一圖形,確保圖形是默認的圖形,並推送名稱范圍和變量范圍。
如果name_or_scope不是None,則使用as is。如果scope是None,則使用default_name。在這種情況下,如果以前在同一范圍內使用過相同的名稱,則通過添加_N來使其具有唯一性。
變量范圍允許您創建新變量並共享已創建的變量,同時提供檢查以防止意外創建或共享。在本文中我們提供了幾個基本示例。
如何創建一個新變量:
with tf.variable_scope("foo"): with tf.variable_scope("bar"): v = tf.get_variable("v", [1]) assert v.name == "foo/bar/v:0"
(6)tf.nn.relu函數
解釋
這個函數的作用是計算激活函數relu,即max(features, 0)。即將矩陣中每行的非最大值置0。
類似的還有tf.sigmoid , tf.tanh
函數定義:
>>> help(tf.nn.relu) Help on function relu in module tensorflow.python.ops.gen_nn_ops: relu(features, name=None) Computes rectified linear: `max(features, 0)`. Args: features: A `Tensor`. Must be one of the following types: `float32`, `float64`, `int32`, `uint8`, `int16`, `int8`, `int64`, `bfloat16`, `uint16`, `half`, `uint32`, `uint64`. name: A name for the operation (optional). Returns: A `Tensor`. Has the same type as `features`.
程序示例:
#!/usr/bin/env python # -*- coding: utf-8 -*- import tensorflow as tf a = tf.constant([-1.0, 2.0]) with tf.Session() as sess: b = tf.nn.relu(a) print sess.run(b)
運行結果:
[0. 2.]
(7)tf.nn.bias_add函數
函數定義:
tf.nn.bias_add(value, bias, data_format=None, name=None) 對value加一偏置量 此函數為tf.add的特殊情況,bias僅為一維, 函數通過廣播機制進行與value求和, 數據格式可以與value不同,返回為與value相同格式
官方釋義:
>>> help(tf.nn.bias_add) Help on function bias_add in module tensorflow.python.ops.nn_ops: bias_add(value, bias, data_format=None, name=None) Adds `bias` to `value`. This is (mostly) a special case of `tf.add` where `bias` is restricted to 1-D. Broadcasting is supported, so `value` may have any number of dimensions. Unlike `tf.add`, the type of `bias` is allowed to differ from `value` in the case where both types are quantized. Args: value: A `Tensor` with type `float`, `double`, `int64`, `int32`, `uint8`, `int16`, `int8`, `complex64`, or `complex128`. bias: A 1-D `Tensor` with size matching the last dimension of `value`. Must be the same type as `value` unless `value` is a quantized type, in which case a different quantized type may be used. data_format: A string. 'NHWC' and 'NCHW' are supported. name: A name for the operation (optional). Returns: A `Tensor` with the same type as `value`.
使用示例:
out = tf.nn.bias_add(mergeFeatureMap, b)
(8)tf.nn.xw_plus_b函數
官方解釋:
>>> help(tf.nn.xw_plus_b) Help on function xw_plus_b in module tensorflow.python.ops.nn_ops: xw_plus_b(x, weights, biases, name=None) Computes matmul(x, weights) + biases. Args: x: a 2D tensor. Dimensions typically: batch, in_units weights: a 2D tensor. Dimensions typically: in_units, out_units biases: a 1D tensor. Dimensions: out_units name: A name for the operation (optional). If not specified "xw_plus_b" is used. Returns: A 2-D Tensor computing matmul(x, weights) + biases. Dimensions typically: batch, out_units.
使用示例:
out = tf.nn.xw_plus_b(x, w, b, name = scope.name)
解釋:
xw_plus_b(x, weights, biases, name=None)相當於matmul(x, weights) + biases.
(9)tf.nn.conv2d函數
官方解釋:
>>> help(tf.nn.conv2d) Help on function conv2d in module tensorflow.python.ops.gen_nn_ops: conv2d(input, filter, strides, padding, use_cudnn_on_gpu=True, data_format='NHWC', dilations=[1, 1, 1, 1], name=None) Computes a 2-D convolution given 4-D `input` and `filter` tensors. Given an input tensor of shape `[batch, in_height, in_width, in_channels]` and a filter / kernel tensor of shape `[filter_height, filter_width, in_channels, out_channels]`, this op performs the following: 1. Flattens the filter to a 2-D matrix with shape `[filter_height * filter_width * in_channels, output_channels]`. 2. Extracts image patches from the input tensor to form a *virtual* tensor of shape `[batch, out_height, out_width, filter_height * filter_width * in_channels]`. 3. For each patch, right-multiplies the filter matrix and the image patch vector. In detail, with the default NHWC format, output[b, i, j, k] = sum_{di, dj, q} input[b, strides[1] * i + di, strides[2] * j + dj, q] * filter[di, dj, q, k] Must have `strides[0] = strides[3] = 1`. For the most common case of the same horizontal and vertices strides, `strides = [1, stride, stride, 1]`. Args: input: A `Tensor`. Must be one of the following types: `half`, `bfloat16`, `float32`. A 4-D tensor. The dimension order is interpreted according to the value of `data_format`, see below for details. filter: A `Tensor`. Must have the same type as `input`. A 4-D tensor of shape `[filter_height, filter_width, in_channels, out_channels]` strides: A list of `ints`. 1-D tensor of length 4. The stride of the sliding window for each dimension of `input`. The dimension order is determined by the value of `data_format`, see below for details. padding: A `string` from: `"SAME", "VALID"`. The type of padding algorithm to use. use_cudnn_on_gpu: An optional `bool`. Defaults to `True`. data_format: An optional `string` from: `"NHWC", "NCHW"`. Defaults to `"NHWC"`. Specify the data format of the input and output data. With the default format "NHWC", the data is stored in the order of: [batch, height, width, channels]. Alternatively, the format could be "NCHW", the data storage order of: [batch, channels, height, width]. dilations: An optional list of `ints`. Defaults to `[1, 1, 1, 1]`. 1-D tensor of length 4. The dilation factor for each dimension of `input`. If set to k > 1, there will be k-1 skipped cells between each filter element on that dimension. The dimension order is determined by the value of `data_format`, see above for details. Dilations in the batch and depth dimensions must be 1. name: A name for the operation (optional). Returns: A `Tensor`. Has the same type as `input`.
函數釋義:
此函數的作用是在給定四維輸入(input)和權重W(filter)的情況下計算二維卷積。 參數解釋: input: 一個Tensor,每個元素的格式必須為float32或float64. input的形狀:[batch,in_height,in_width,in_channels], batch為訓練過程中每迭代一次迭代的照片數。 in_height,in_width分別為圖片的高和寬 in_channels為圖片的道。 filter: 一個Tensor,每個元素的類型和input類型一致。 filter的形狀:[filter_height,filter_width,in_channels,out_channels] 分別為權重的height,width,輸入的channels和輸出的channels stride: 長度為4的list,元素類型為int。表示每一維度滑動的步長。 需要注意的是,strides[0]=strides[3]=1. padding: 可選參數為"Same","VALID" 邊距,一般設為0,即padding='SAME' use_cudnn_on_gpu: bool類型,有True和False兩種選擇。 name: 此操作的名字
函數執行以下操作: 1.將參數filter變為一個二維矩陣,形狀為:[filter_height*filter_width*in_channels,output_channels] 2.將輸入(input)轉化為一個具有如下形狀的Tensor,形狀為:[batch,out_height,out_width,filter_height * filter_width * in_channels] 3.將filter矩陣和步驟2得到的矩陣相乘。
函數的返回值:
元素類型和input相同。
output[b, i, j, k] =sum_{di, dj, q} input[b, strides[1] * i + di, strides[2] * j + dj, q] * filter[di, dj, q, k]
編程示例:
kernel = tf.Variable(tf.truncated_normal([3,3,384,256], dtype=tf.float32, stddev=1e-1), name='weights') conv = tf.nn.conv2d(conv3, kernel, [1,1,1,1],padding='SAME')
(10)tf.constant函數
官方解釋:
>>> help(tf.constant) Help on function constant in module tensorflow.python.framework.constant_op: constant(value, dtype=None, shape=None, name='Const', verify_shape=False) Creates a constant tensor. The resulting tensor is populated with values of type `dtype`, as specified by arguments `value` and (optionally) `shape` (see examples below). The argument `value` can be a constant value, or a list of values of type `dtype`. If `value` is a list, then the length of the list must be less than or equal to the number of elements implied by the `shape` argument (if specified). In the case where the list length is less than the number of elements specified by `shape`, the last element in the list will be used to fill the remaining entries. The argument `shape` is optional. If present, it specifies the dimensions of the resulting tensor. If not present, the shape of `value` is used. If the argument `dtype` is not specified, then the type is inferred from the type of `value`. For example: ```python # Constant 1-D Tensor populated with value list. tensor = tf.constant([1, 2, 3, 4, 5, 6, 7]) => [1 2 3 4 5 6 7] # Constant 2-D tensor populated with scalar value -1. tensor = tf.constant(-1.0, shape=[2, 3]) => [[-1. -1. -1.] [-1. -1. -1.]] ``` Args: value: A constant value (or list) of output type `dtype`. dtype: The type of the elements of the resulting tensor. shape: Optional dimensions of resulting tensor. name: Optional name for the tensor. verify_shape: Boolean that enables verification of a shape of values. Returns: A Constant Tensor. Raises: TypeError: if shape is incorrectly specified or unsupported.
程序示例:
https://blog.csdn.net/qq_26591517/article/details/80198471