Cam(Class Activation Mapping)是一個(gè)很有意思的算法，他能夠?qū)⑸窠?jīng)網(wǎng)絡(luò)到底在關(guān)注什么可視化的表現(xiàn)出來。但同時(shí)它的實(shí)現(xiàn)卻又如此簡介，相比NIN，googLenet這些使用GAP（Global Average Pooling）用來代替全連接層，他卻將其輸出的權(quán)重和featuremap相乘，累加，將其用圖像表示出來。
其網(wǎng)絡(luò)架構(gòu)如下

Class Activation Mapping具體論文

當(dāng)然Cam的目的并不僅僅是將其表示出來，神經(jīng)網(wǎng)絡(luò)所關(guān)注的地方，通常就是物體所在的地方，因此它可以輔助訓(xùn)練檢測網(wǎng)絡(luò)。
因此就有了PlacesNet。

論文

在這里可以體驗(yàn)

網(wǎng)絡(luò)上基本都是基于AlexNet等網(wǎng)絡(luò)，其實(shí)任何網(wǎng)絡(luò)，只要加一層全局池化層就可以幫助我們將CNN關(guān)注什么表示出來，因此我對(duì)Tensorflow官方Mnist的CNN網(wǎng)絡(luò)進(jìn)行少量的修改，實(shí)現(xiàn)了CAM。只是將最后的全連接層，改為了全局池化層。

CAM的核心公式很簡單
\[ S_c = \sum_{k} {W_k^c} {\sum_{x,y} {f_k(x,y)}} \]

將全局平均池化層輸出的權(quán)重乘上feature map累加

import tensorflow as tf from tensorflow.examples.tutorials.mnist import input_data import matplotlib.pyplot as plt import numpy as np %matplotlib inlinemnist = input_data.read_data_sets("/tmp/data/", one_hot=True) /home/lyn/anaconda3/lib/python3.6/importlib/_bootstrap.py:205: RuntimeWarning: compiletime version 3.5 of module 'tensorflow.python.framework.fast_tensor_util' does not match runtime version 3.6return f(*args, **kwds)Extracting /tmp/data/train-images-idx3-ubyte.gz Extracting /tmp/data/train-labels-idx1-ubyte.gz Extracting /tmp/data/t10k-images-idx3-ubyte.gz Extracting /tmp/data/t10k-labels-idx1-ubyte.gz

在讀入數(shù)據(jù)后，設(shè)定基本的學(xué)習(xí)參數(shù)

# Training Parameters learning_rate = 0.001 num_steps = 10000 batch_size = 128 display_step = 10# Network Parameters num_input = 784 # MNIST data input (img shape: 28*28) num_classes = 10 # MNIST total classes (0-9 digits)# tf Graph input X = tf.placeholder(tf.float32, [None, num_input]) Y = tf.placeholder(tf.int32, [None, num_classes]) # Create some wrappers for simplicity def conv2d(x, W, b, strides=1):# Conv2D wrapper, with bias and relu activationx = tf.nn.conv2d(x, W, strides=[1, strides, strides, 1], padding='SAME')x = tf.nn.bias_add(x, b)return tf.nn.relu(x)def maxpool2d(x, k=2):# MaxPool2D wrapperreturn tf.nn.max_pool(x, ksize=[1, k, k, 1], strides=[1, k, k, 1],padding='SAME')

在實(shí)際使用中我們需要獲得得feature map，與全局池化層相乘并累加

def conv_layers(x,weights,biases):x = tf.reshape(x, shape=[-1, 28, 28, 1])# Convolution Layerconv1 = conv2d(x, weights['wc1'], biases['bc1'])# Max Pooling (down-sampling)conv1 = maxpool2d(conv1, k=2)# Convolution Layerconv2 = conv2d(conv1, weights['wc2'], biases['bc2'])# Max Pooling (down-sampling)conv2 = maxpool2d(conv2, k=2)return conv2def out_layer(conv2,weights,biases):gap = tf.nn.avg_pool(conv2,ksize=[1,7,7,1],strides=[1,7,7,1],padding="SAME")gap = tf.reshape(gap,[-1,128])out = tf.add(tf.matmul(gap, weights['out']), biases['out'])return outdef generate_heatmap(conv2,label,weights):conv2_resized = tf.image.resize_images(conv2,[28,28])label_w = tf.gather(tf.transpose(weights['out']),label)label_w = tf.reshape(label_w,[-1,128,1])conv2_resized = tf.reshape(conv2_resized,[-1,28*28,128])classmap = tf.matmul( conv2_resized, label_w )classmap = tf.reshape( classmap, [-1, 28,28] )return classmap # Store layers weight & bias weights = {# 5x5 conv, 1 input, 32 outputs'wc1': tf.Variable(tf.random_normal([5, 5, 1, 32])),# 5x5 conv, 32 inputs, 64 outputs'wc2': tf.Variable(tf.random_normal([5, 5, 32, 128])),'out': tf.Variable(tf.random_normal([128, num_classes])) }biases = {'bc1': tf.Variable(tf.random_normal([32])),'bc2': tf.Variable(tf.random_normal([128])),'bd1': tf.Variable(tf.random_normal([1024])),'out': tf.Variable(tf.random_normal([num_classes])) }# Construct model conv2 = conv_layers(X, weights, biases) logits = out_layer(conv2,weights, biases) prediction = tf.nn.softmax(logits) classmap = generate_heatmap(conv2,tf.argmax(prediction,1),weights) loss_op = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits, labels=Y)) optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate) train_op = optimizer.minimize(loss_op)# Evaluate model correct_pred = tf.equal(tf.argmax(prediction, 1), tf.argmax(Y, 1)) accuracy = tf.reduce_mean(tf.cast(correct_pred, tf.float32))# Initialize the variables (i.e. assign their default value) init = tf.global_variables_initializer() sess = tf.Session() sess.run(init)for step in range(1, num_steps+1):batch_x, batch_y = mnist.train.next_batch(batch_size)# Run optimization op (backprop)sess.run(train_op, feed_dict={X: batch_x, Y: batch_y})if step % display_step == 0 or step == 1:# Calculate batch loss and accuracyloss, acc = sess.run([loss_op, accuracy], feed_dict={X: batch_x,Y: batch_y})print("Step " + str(step) + ", Minibatch Loss= " + \"{:.4f}".format(loss) + ", Training Accuracy= " + \"{:.3f}".format(acc))print("Optimization Finished!")# Calculate accuracy for 256 MNIST test images print("Testing Accuracy:", \sess.run(accuracy, feed_dict={X: mnist.test.images[:256],Y: mnist.test.labels[:256],})) sess.run(conv2,feed_dict={X:mnist.test.images[:10],Y:mnist.test.labels[:10]}) classmaps = sess.run(classmap,feed_dict={X:mnist.test.images[:10],Y:mnist.test.labels[:10]}) Testing Accuracy: 0.976562 for i in range(9):plt.subplot(33*10+i+1)plt.axis("off")plt.imshow(classmaps[i],cmap="gray")plt.title("label is"+str(np.argmax(mnist.test.labels[i])))

黑色為正權(quán)重點(diǎn)，白色為負(fù)權(quán)重點(diǎn)。

當(dāng)然由于網(wǎng)絡(luò)太淺，使用了全局平均池化以后，訓(xùn)練時(shí)間大大增長，準(zhǔn)確率也不如之前。

轉(zhuǎn)載于:https://www.cnblogs.com/lynsyklate/p/7966171.html

總結(jié)

以上是生活随笔為你收集整理的在CNN上增加一层CAM告诉你CNN到底关注什么的全部內(nèi)容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網(wǎng)站內(nèi)容還不錯(cuò)，歡迎將生活随笔推薦給好友。