classify CIFAR Image With CNN

本教程演示如何训练一个简单的卷积神经网络(CNN)来对CIFAR图像进行分类。因为本教程使用Keras Sequential API，所以创建和训练我们的模型只需要几行代码

1.1、加载数据

def loadData():
    # 加载数据，cifar数据集 有10个分类，每个分类有6000张彩色图像
    (train_images, train_labels), (test_images, test_labels) = datasets.cifar10.load_data()

    # 归一化
    train_images, test_images = train_images / 255.0, test_images / 255.0
    return train_labels, test_images

1.1.1、解决下载cifar数据集比较慢的问题

下载一份cifar数据集放到~\.keras\datasets目录下，
然后修改cifar10.py, 点击datasets.cifar10.load_data()找到cifar10.py文件

@keras_export('keras.datasets.cifar10.load_data')
def load_data():
  """Loads CIFAR10 dataset.

  Returns:
      Tuple of Numpy arrays: `(x_train, y_train), (x_test, y_test)`.
  """
  dirname = 'cifar-10-batches-py'
  origin = 'https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz'
  path = get_file(
      dirname,
      origin=origin,
      untar=True,   
      # 此处注释掉，防止验证文件hash
      # file_hash= '6d958be074577803d12ecdefd02955f39262c83c16fe9348329d7fe0b5c001ce'
  )
  ......
}

1.2、验证数据

def verifyData(train_images, train_labels):
    # 查看前25张图片，验证数据集的正确性
    class_names = ['airplane', 'automobile', 'bird', 'cat', 'deer',
                   'dog', 'frog', 'horse', 'ship', 'truck']

    plt.figure(figsize=(10,10))
    for i in range(25):
        plt.subplot(5,5,i+1)
        plt.xticks([])
        plt.yticks([])
        plt.grid(False)
        plt.imshow(train_images[i], cmap=plt.cm.binary)
        # The CIFAR labels happen to be arrays,
        # which is why you need the extra index
        plt.xlabel(class_names[train_labels[i][0]])
    plt.show()

1.3、创建卷积基

1.3.1、卷积层

# 创建卷积基
model = models.Sequential()
model.add(layers.Conv2D(32, (3, 3), activation='relu', input_shape=(32, 32, 3)))

1.3.2、池化层

	model.add(layers.MaxPooling2D((2, 2)))

1.3.3、全连接层

# 全连接层
model.add(layers.Flatten())
model.add(layers.Dense(64, activation='relu'))

完整的构建模型

train_images, train_labels, test_images, test_labels = loadData()
print(train_images.shape)
# verifyData(train_images, train_labels)

# 创建卷积基
model = models.Sequential()
model.add(layers.Conv2D(32, (3, 3), activation='relu', input_shape=(32, 32, 3)))
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Conv2D(64, (3, 3), activation='relu'))
model.add(layers.MaxPooling2D((2, 2)))
model.add(layers.Conv2D(64, (3, 3), activation='relu'))

# 全连接层
model.add(layers.Flatten())
model.add(layers.Dense(64, activation='relu'))
# 分类
model.add(layers.Dense(10, activation='softmax')) # 输出层：10个分类， 使用softmax激活函数



print(model.summary())

Model: "sequential"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
conv2d (Conv2D)              (None, 30, 30, 32)        896       
_________________________________________________________________
max_pooling2d (MaxPooling2D) (None, 15, 15, 32)        0         
_________________________________________________________________
conv2d_1 (Conv2D)            (None, 13, 13, 64)        18496     
_________________________________________________________________
max_pooling2d_1 (MaxPooling2 (None, 6, 6, 64)          0         
_________________________________________________________________
conv2d_2 (Conv2D)            (None, 4, 4, 64)          36928     
_________________________________________________________________
flatten (Flatten)            (None, 1024)              0         
_________________________________________________________________
dense (Dense)                (None, 64)                65600     
_________________________________________________________________
dense_1 (Dense)              (None, 10)                650       
=================================================================
Total params: 122,570
Trainable params: 122,570
Non-trainable params: 0
_________________________________________________________________
None

1.4、编译模型

# 编译模型
model.compile(optimizer='adam',
              loss='sparse_categorical_crossentropy',
              metrics=['accuracy'])

1.5、训练模型

history = model.fit(train_images, train_labels, epochs=10,
                    validation_data=(test_images, test_labels))

49472/50000 [============================>.] - ETA: 0s - loss: 0.6245 - accuracy: 0.7791
49728/50000 [============================>.] - ETA: 0s - loss: 0.6242 - accuracy: 0.7792
49984/50000 [============================>.] - ETA: 0s - loss: 0.6248 - accuracy: 0.7789
50000/50000 [==============================] - 12s 249us/sample - loss: 0.6247 - accuracy: 0.7790 - val_loss: 0.8589 - val_accuracy: 0.7174

cudnn未安装

tensorflow.python.framework.errors_impl.UnknownError:  Failed to get convolution algorithm. 
This is probably because cuDNN failed to initialize, so try looking to see if a warning log message was printed above.
  [[node sequential/conv2d/Conv2D (defined at \setup\Python37\lib\site-packages\tensorflow_core\python\framework\ops.py:1751) ]] [Op:__inference_distributed_function_1055]

1.6、评估模型

plt.plot(history.history['accuracy'], label='accuracy')
plt.plot(history.history['val_accuracy'], label = 'val_accuracy')
plt.xlabel('Epoch')
plt.ylabel('Accuracy')
plt.ylim([0.5, 1])
plt.legend(loc='lower right')
plt.show()
test_loss, test_acc = model.evaluate(test_images,  test_labels, verbose=2)

10000/1 - 1s - loss: 0.7876 - accuracy: 0.7174