03-CNN服装图像分类

import tensorflow as tf
from tensorflow.keras import datasets,layers,models
import matplotlib.pyplot as plt

#数据导入
(train_images,train_labels),\
(test_images,test_labels) = datasets.fashion_mnist.load_data()

#数据探索
plt.figure()
plt.imshow(train_images[0])
plt.colorbar()
plt.grid(False)
plt.show()

#归一化
train_images,test_images = train_images/255.0,test_images/255.0
train_images.shape,test_images.shape,train_labels.shape,test_labels.shape

((60000, 28, 28), (10000, 28, 28), (60000,), (10000,))

#调整图片格式
train_images = train_images.reshape((60000,28,28,1))
test_images = test_images.reshape((10000,28,28,1))

train_images.shape,test_images.shape,train_labels.shape,test_labels.shape

((60000, 28, 28, 1), (10000, 28, 28, 1), (60000,), (10000,))

#可视化
class_names = ['T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat',
               'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle boot']

plt.figure(figsize=(20,10))
for i in range(20):
    plt.subplot(5,10,i+1)
    plt.xticks([])
    plt.yticks([])
    plt.grid(False)
    plt.imshow(train_images[i],cmap=plt.cm.binary)
    plt.xlabel(class_names[train_labels[i]])
plt.show()

#构建CNN网络
model = models.Sequential([
    layers.Conv2D(32,(3,3),activation='relu',input_shape=(28,28,1)),#input_shape=(28,28,1)传入灰度图像，卷积层1，卷积核3*3
    layers.MaxPool2D((2,2)),#池化层1，2*2采样
    layers.Conv2D(64,(3,3),activation='relu'),#卷积层2，卷积核3*3
    layers.MaxPool2D(2,2),#池化层2，2*2采样
    layers.Conv2D(64,(3,3),activation='relu'),#卷积层3，卷积核3*3
    
    layers.Flatten(),#Flatern层，连接卷积层与全连接层
    layers.Dense(64,activation='relu'),#全连接层，特征进一步提取
    layers.Dense(10)#输出层，输出预期结果
])
model.summary()

Model: "sequential_1"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
conv2d_3 (Conv2D)            (None, 26, 26, 32)        320       
_________________________________________________________________
max_pooling2d_2 (MaxPooling2 (None, 13, 13, 32)        0         
_________________________________________________________________
conv2d_4 (Conv2D)            (None, 11, 11, 64)        18496     
_________________________________________________________________
max_pooling2d_3 (MaxPooling2 (None, 5, 5, 64)          0         
_________________________________________________________________
conv2d_5 (Conv2D)            (None, 3, 3, 64)          36928     
_________________________________________________________________
flatten_1 (Flatten)          (None, 576)               0         
_________________________________________________________________
dense_2 (Dense)              (None, 64)                36928     
_________________________________________________________________
dense_3 (Dense)              (None, 10)                650       
=================================================================
Total params: 93,322
Trainable params: 93,322
Non-trainable params: 0
_________________________________________________________________

#编译
model.compile(optimizer='adam',#优化器
             loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),#损失函数
             metrics=['accuracy'])#指定监控指标

#训练模型
history = model.fit(train_images,train_labels,epochs=10,
                   validation_data=(test_images,test_labels))

Train on 60000 samples, validate on 10000 samples
Epoch 1/10
60000/60000 [==============================] - 144s 2ms/sample - loss: 0.4953 - accuracy: 0.8195 - val_loss: 0.3678 - val_accuracy: 0.8666
Epoch 2/10
60000/60000 [==============================] - 140s 2ms/sample - loss: 0.3212 - accuracy: 0.8831 - val_loss: 0.3217 - val_accuracy: 0.8832
Epoch 3/10
60000/60000 [==============================] - 140s 2ms/sample - loss: 0.2728 - accuracy: 0.8999 - val_loss: 0.2854 - val_accuracy: 0.8988
Epoch 4/10
60000/60000 [==============================] - 139s 2ms/sample - loss: 0.2427 - accuracy: 0.9099 - val_loss: 0.2898 - val_accuracy: 0.8928
Epoch 5/10
60000/60000 [==============================] - 140s 2ms/sample - loss: 0.2200 - accuracy: 0.9198 - val_loss: 0.2645 - val_accuracy: 0.9049
Epoch 6/10
60000/60000 [==============================] - 139s 2ms/sample - loss: 0.1992 - accuracy: 0.9268 - val_loss: 0.2607 - val_accuracy: 0.9043
Epoch 7/10
60000/60000 [==============================] - 137s 2ms/sample - loss: 0.1835 - accuracy: 0.9319 - val_loss: 0.2475 - val_accuracy: 0.9114
Epoch 8/10
60000/60000 [==============================] - 139s 2ms/sample - loss: 0.1686 - accuracy: 0.9369 - val_loss: 0.2867 - val_accuracy: 0.9043
Epoch 9/10
60000/60000 [==============================] - 138s 2ms/sample - loss: 0.1555 - accuracy: 0.9416 - val_loss: 0.2651 - val_accuracy: 0.9104
Epoch 10/10
60000/60000 [==============================] - 135s 2ms/sample - loss: 0.1423 - accuracy: 0.9464 - val_loss: 0.2734 - val_accuracy: 0.9090

#预测
plt.imshow(test_images[1])

<matplotlib.image.AxesImage at 0x7fda29d3e990>

import numpy as np
pre = model.predict(test_images)
print(class_names[np.argmax(pre[1])])

Pullover

#模型评估
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 - 10s - loss: 0.1572 - accuracy: 0.9090

print('测试准确度为',test_acc)

测试准确度为 0.909