- 🍨 本文为🔗365天深度学习训练营 中的学习记录博客
- 🍖 原作者:K同学啊
前期准备
import tensorflow as tf
gpus = tf.config.list_physical_devices("GPU")
if gpus:
tf.config.experimental.set_memory_growth(gpus[0], True) #设置GPU显存用量按需使用
tf.config.set_visible_devices([gpus[0]],"GPU")
gpus
[]
from tensorflow import keras
from tensorflow.keras import layers,models
import numpy as np
import matplotlib.pyplot as plt
import os,PIL,pathlib
data_dir = r"C:\Users\11054\Desktop\kLearning\t7_learning\data"
data_dir = pathlib.Path(data_dir)
image_count = len(list(data_dir.glob('*/*.png')))
print("图片总数为:",image_count)
图片总数为: 1200
数据预处理
batch_size = 32
img_height = 224
img_width = 224
"""
关于image_dataset_from_directory()的详细介绍可以参考文章:https://mtyjkh.blog.csdn.net/article/details/117018789
"""
train_ds = tf.keras.preprocessing.image_dataset_from_directory(
data_dir,
validation_split=0.2,
subset="training",
seed=123,
image_size=(img_height, img_width),
batch_size=batch_size)
"""
关于image_dataset_from_directory()的详细介绍可以参考文章:https://mtyjkh.blog.csdn.net/article/details/117018789
"""
val_ds = tf.keras.preprocessing.image_dataset_from_directory(
data_dir,
validation_split=0.2,
subset="validation",
seed=123,
image_size=(img_height, img_width),
batch_size=batch_size)
class_names = train_ds.class_names
print(class_names)
Found 1200 files belonging to 4 classes.
Using 960 files for training.
Found 1200 files belonging to 4 classes.
Using 240 files for validation.
['Dark', 'Green', 'Light', 'Medium']
可视化数据
plt.figure(figsize=(10, 4)) # 图形的宽为10高为5
for images, labels in train_ds.take(1):
for i in range(10):
ax = plt.subplot(2, 5, i + 1)
plt.imshow(images[i].numpy().astype("uint8"))
plt.title(class_names[labels[i]])
plt.axis("off")
for image_batch, labels_batch in train_ds:
print(image_batch.shape)
print(labels_batch.shape)
break
(32, 224, 224, 3)
(32,)
配置数据集
AUTOTUNE = tf.data.AUTOTUNE
train_ds = train_ds.cache().shuffle(1000).prefetch(buffer_size=AUTOTUNE)
val_ds = val_ds.cache().prefetch(buffer_size=AUTOTUNE)
normalization_layer = layers.Rescaling(1./255)
train_ds = train_ds.map(lambda x, y: (normalization_layer(x), y))
val_ds = val_ds.map(lambda x, y: (normalization_layer(x), y))
image_batch, labels_batch = next(iter(val_ds))
first_image = image_batch[0]
# 查看归一化后的数据
print(np.min(first_image), np.max(first_image))
0.0 1.0
构建VGG-16网络
model = tf.keras.applications.VGG16(weights='imagenet')
# from tensorflow.keras import layers, models, Input
# from tensorflow.keras.models import Model
# from tensorflow.keras.layers import Conv2D, MaxPooling2D, Dense, Flatten, Dropout
#
# def VGG16(nb_classes, input_shape):
# input_tensor = Input(shape=input_shape)
# # 1st block
# x = Conv2D(64, (3,3), activation='relu', padding='same',name='block1_conv1')(input_tensor)
# x = Conv2D(64, (3,3), activation='relu', padding='same',name='block1_conv2')(x)
# x = MaxPooling2D((2,2), strides=(2,2), name = 'block1_pool')(x)
# # 2nd block
# x = Conv2D(128, (3,3), activation='relu', padding='same',name='block2_conv1')(x)
# x = Conv2D(128, (3,3), activation='relu', padding='same',name='block2_conv2')(x)
# x = MaxPooling2D((2,2), strides=(2,2), name = 'block2_pool')(x)
# # 3rd block
# x = Conv2D(256, (3,3), activation='relu', padding='same',name='block3_conv1')(x)
# x = Conv2D(256, (3,3), activation='relu', padding='same',name='block3_conv2')(x)
# x = Conv2D(256, (3,3), activation='relu', padding='same',name='block3_conv3')(x)
# x = MaxPooling2D((2,2), strides=(2,2), name = 'block3_pool')(x)
# # 4th block
# x = Conv2D(512, (3,3), activation='relu', padding='same',name='block4_conv1')(x)
# x = Conv2D(512, (3,3), activation='relu', padding='same',name='block4_conv2')(x)
# x = Conv2D(512, (3,3), activation='relu', padding='same',name='block4_conv3')(x)
# x = MaxPooling2D((2,2), strides=(2,2), name = 'block4_pool')(x)
# # 5th block
# x = Conv2D(512, (3,3), activation='relu', padding='same',name='block5_conv1')(x)
# x = Conv2D(512, (3,3), activation='relu', padding='same',name='block5_conv2')(x)
# x = Conv2D(512, (3,3), activation='relu', padding='same',name='block5_conv3')(x)
# x = MaxPooling2D((2,2), strides=(2,2), name = 'block5_pool')(x)
# # full connection
# x = Flatten()(x)
# x = Dense(4096, activation='relu', name='fc1')(x)
# x = Dense(4096, activation='relu', name='fc2')(x)
# output_tensor = Dense(nb_classes, activation='softmax', name='predictions')(x)
#
# model = Model(input_tensor, output_tensor)
# return model
#
# model=VGG16(len(class_names), (img_width, img_height, 3))
# model.summary()
<Functional name=vgg16, built=True>
网络结构图
编译
# 设置初始学习率
initial_learning_rate = 1e-4
lr_schedule = tf.keras.optimizers.schedules.ExponentialDecay(
initial_learning_rate,
decay_steps=30, # 敲黑板!!!这里是指 steps,不是指epochs
decay_rate=0.92, # lr经过一次衰减就会变成 decay_rate*lr
staircase=True)
# 设置优化器
opt = tf.keras.optimizers.Adam(learning_rate=initial_learning_rate)
model.compile(optimizer=opt,
loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
metrics=['accuracy'])
训练模型
epochs = 20
history = model.fit(
train_ds,
validation_data=val_ds,
epochs=epochs
)
Epoch 1/20
C:\Users\11054\.conda\envs\py311\Lib\site-packages\keras\src\backend\tensorflow\nn.py:602: UserWarning: "`sparse_categorical_crossentropy` received `from_logits=True`, but the `output` argument was produced by a Softmax activation and thus does not represent logits. Was this intended?
output, from_logits = _get_logits(
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m311s[0m 10s/step - accuracy: 0.2452 - loss: 2.3790 - val_accuracy: 0.4708 - val_loss: 1.1447
Epoch 2/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m322s[0m 10s/step - accuracy: 0.5306 - loss: 0.9733 - val_accuracy: 0.8000 - val_loss: 0.6850
Epoch 3/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m305s[0m 10s/step - accuracy: 0.8199 - loss: 0.4469 - val_accuracy: 0.8792 - val_loss: 0.3026
Epoch 4/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m305s[0m 10s/step - accuracy: 0.9060 - loss: 0.2226 - val_accuracy: 0.9125 - val_loss: 0.2860
Epoch 5/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m305s[0m 10s/step - accuracy: 0.9341 - loss: 0.1610 - val_accuracy: 0.8417 - val_loss: 0.4898
Epoch 6/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m305s[0m 10s/step - accuracy: 0.9024 - loss: 0.2810 - val_accuracy: 0.9333 - val_loss: 0.1948
Epoch 7/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m323s[0m 10s/step - accuracy: 0.9804 - loss: 0.0554 - val_accuracy: 0.9625 - val_loss: 0.1087
Epoch 8/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m305s[0m 10s/step - accuracy: 0.9862 - loss: 0.0477 - val_accuracy: 0.9875 - val_loss: 0.0366
Epoch 9/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m307s[0m 10s/step - accuracy: 0.9967 - loss: 0.0159 - val_accuracy: 0.9958 - val_loss: 0.0105
Epoch 10/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m321s[0m 10s/step - accuracy: 0.9995 - loss: 0.0034 - val_accuracy: 0.9958 - val_loss: 0.0112
Epoch 11/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m304s[0m 10s/step - accuracy: 1.0000 - loss: 9.6800e-04 - val_accuracy: 0.9958 - val_loss: 0.0102
Epoch 12/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m323s[0m 10s/step - accuracy: 1.0000 - loss: 2.3597e-04 - val_accuracy: 0.9958 - val_loss: 0.0092
Epoch 13/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m306s[0m 10s/step - accuracy: 1.0000 - loss: 1.6429e-04 - val_accuracy: 0.9958 - val_loss: 0.0114
Epoch 14/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m305s[0m 10s/step - accuracy: 1.0000 - loss: 1.2267e-04 - val_accuracy: 0.9917 - val_loss: 0.0120
Epoch 15/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m304s[0m 10s/step - accuracy: 1.0000 - loss: 5.3608e-05 - val_accuracy: 0.9958 - val_loss: 0.0130
Epoch 16/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m305s[0m 10s/step - accuracy: 1.0000 - loss: 2.8213e-05 - val_accuracy: 0.9917 - val_loss: 0.0137
Epoch 17/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m308s[0m 10s/step - accuracy: 1.0000 - loss: 1.5108e-05 - val_accuracy: 0.9917 - val_loss: 0.0146
Epoch 18/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m320s[0m 10s/step - accuracy: 1.0000 - loss: 1.2032e-05 - val_accuracy: 0.9958 - val_loss: 0.0145
Epoch 19/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m308s[0m 10s/step - accuracy: 1.0000 - loss: 1.0658e-05 - val_accuracy: 0.9958 - val_loss: 0.0153
Epoch 20/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m308s[0m 10s/step - accuracy: 1.0000 - loss: 7.8087e-06 - val_accuracy: 0.9958 - val_loss: 0.0147
acc = history.history['accuracy']
val_acc = history.history['val_accuracy']
loss = history.history['loss']
val_loss = history.history['val_loss']
epochs_range = range(epochs)
plt.figure(figsize=(12, 4))
plt.subplot(1, 2, 1)
plt.plot(epochs_range, acc, label='Training Accuracy')
plt.plot(epochs_range, val_acc, label='Validation Accuracy')
plt.legend(loc='lower right')
plt.title('Training and Validation Accuracy')
plt.subplot(1, 2, 2)
plt.plot(epochs_range, loss, label='Training Loss')
plt.plot(epochs_range, val_loss, label='Validation Loss')
plt.legend(loc='upper right')
plt.title('Training and Validation Loss')
plt.show()
#个人总结
- 掌握了VGG16 TensorFlow的调用