本文已加入 🚀 Python AI 计划，从一个Python小白到一个AI大神，你所需要的所有知识都在 这里 了。
数据中一共包含1462张数据图片，分为浮标、游轮、渡船、货船、贡多拉、充气船、皮划艇、纸船、帆船等9类。我们将通过ResNet50算法实现这9类目标的识别，最后的准确率为87.0%。
🥇 需要 项目定制、毕设辅导 的同学可以加我V.信：mtyjkh_
我的环境：
- 语言环境：Python3.8
- 编译器：Jupyter lab
- 深度学习环境：TensorFlow2.4.1
我们的代码流程图如下所示：
一、设置GPU
import tensorflow as tf
gpus = tf.config.list_physical_devices("GPU")
if gpus:
gpu0 = gpus[0] #如果有多个GPU，仅使用第0个GPU
tf.config.experimental.set_memory_growth(gpu0, True) #设置GPU显存用量按需使用
tf.config.set_visible_devices([gpu0],"GPU")
import matplotlib.pyplot as plt
import os,PIL,pathlib
import numpy as np
import pandas as pd
import warnings
from tensorflow import keras
warnings.filterwarnings("ignore")#忽略警告信息
plt.rcParams['font.sans-serif'] = ['SimHei'] # 用来正常显示中文标签
plt.rcParams['axes.unicode_minus'] = False # 用来正常显示负号
二、导入数据
1. 导入数据
import pathlib
data_dir = "./29-data/"
data_dir = pathlib.Path(data_dir)
image_count = len(list(data_dir.glob('*/*')))
print("图片总数为：",image_count)
图片总数为： 1462
batch_size = 16
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
通过该方法导入数据时，会同时打乱数据
"""
train_ds = tf.keras.preprocessing.image_dataset_from_directory(
data_dir,
validation_split=0.1,
subset="training",
seed=12,
image_size=(img_height, img_width),
batch_size=batch_size)
Found 1462 files belonging to 9 classes.
Using 1316 files for training.
"""
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
通过该方法导入数据时，会同时打乱数据
"""
val_ds = tf.keras.preprocessing.image_dataset_from_directory(
data_dir,
validation_split=0.1,
subset="validation",
seed=12,
image_size=(img_height, img_width),
batch_size=batch_size)
Found 1462 files belonging to 9 classes.
Using 146 files for validation.
class_names = train_ds.class_names
print("数据类别有：",class_names)
print("需要识别的船一共有%d类"%len(class_names))
数据类别有： ['buoy', 'cruise ship', 'ferry boat', 'freight boat', 'gondola', 'inflatable boat', 'kayak', 'paper boat', 'sailboat']
需要识别的船一共有9类
2. 检查数据
for image_batch, labels_batch in train_ds:
print(image_batch.shape)
print(labels_batch.shape)
break
(16, 224, 224, 3)
(16,)
3. 配置数据集
- shuffle() ： 打乱数据。
- prefetch() ： 预取数据，加速运行，其详细介绍可以参考我前两篇文章，里面都有讲解。
- cache() ： 将数据集缓存到内存当中，加速运行
AUTOTUNE = tf.data.AUTOTUNE
def train_preprocessing(image,label):
return (image/255.0,label)
train_ds = (
train_ds.cache()
.map(train_preprocessing) # 这里可以设置预处理函数
.prefetch(buffer_size=AUTOTUNE)
)
val_ds = (
val_ds.cache()
.map(train_preprocessing) # 这里可以设置预处理函数
.prefetch(buffer_size=AUTOTUNE)
)
4. 数据可视化
plt.figure(figsize=(10, 8)) # 图形的宽为10高为5
plt.suptitle("数据展示")
for images, labels in train_ds.take(1):
for i in range(15):
plt.subplot(4, 5, i + 1)
plt.xticks([])
plt.yticks([])
plt.grid(False)
# 显示图片
plt.imshow(images[i])
# 显示标签
plt.xlabel(class_names[labels[i]-1])
plt.show()
三、构建模型
在这次训练的过程中我发现一个有趣的现象：当我使用复杂的网络时，训练效果不是很理想；当采用相对简单的网络时，效果反而还不错。
from tensorflow.keras import layers, models, Input
from tensorflow.keras.models import Model
from tensorflow.keras.layers import Conv2D, MaxPooling2D, Dense, Flatten, Dropout,BatchNormalization,Activation
# 加载预训练模型
base_model = keras.applications.ResNet50(weights='imagenet', include_top=False, input_shape=(img_width,img_height,3))
for layer in base_model.layers:
layer.trainable = True
# Add layers at the end
X = base_model.output
X = Flatten()(X)
X = Dense(512, kernel_initializer='he_uniform')(X)
X = Dropout(0.5)(X)
X = BatchNormalization()(X)
X = Activation('relu')(X)
X = Dense(16, kernel_initializer='he_uniform')(X)
X = Dropout(0.5)(X)
X = BatchNormalization()(X)
X = Activation('relu')(X)
output = Dense(len(class_names), activation='softmax')(X)
model = Model(inputs=base_model.input, outputs=output)
四、编译
optimizer = tf.keras.optimizers.Adam(lr=1e-4)
model.compile(optimizer=optimizer,
loss='sparse_categorical_crossentropy',
metrics=['accuracy'])
五、训练模型
from tensorflow.keras.callbacks import ModelCheckpoint, Callback, EarlyStopping, ReduceLROnPlateau, LearningRateScheduler
NO_EPOCHS = 50
PATIENCE = 5
VERBOSE = 1
# 设置动态学习率
# annealer = LearningRateScheduler(lambda x: 1e-3 * 0.99 ** (x+NO_EPOCHS))
# 设置早停
earlystopper = EarlyStopping(monitor='loss', patience=PATIENCE, verbose=VERBOSE)
#
checkpointer = ModelCheckpoint('best_model.h5',
monitor='val_accuracy',
verbose=VERBOSE,
save_best_only=True,
save_weights_only=True)
train_model = model.fit(train_ds,
epochs=NO_EPOCHS,
verbose=1,
validation_data=val_ds,
callbacks=[earlystopper, checkpointer])
Epoch 1/50
83/83 [==============================] - 17s 109ms/step - loss: 1.8596 - accuracy: 0.3625 - val_loss: 2.2435 - val_accuracy: 0.3699
Epoch 00001: val_accuracy improved from -inf to 0.36986, saving model to best_model.h5
Epoch 2/50
83/83 [==============================] - 8s 94ms/step - loss: 1.5476 - accuracy: 0.5190 - val_loss: 2.1825 - val_accuracy: 0.1575
......
Epoch 00049: val_accuracy did not improve from 0.90411
Epoch 50/50
83/83 [==============================] - 7s 81ms/step - loss: 0.4809 - accuracy: 0.9111 - val_loss: 0.5607 - val_accuracy: 0.8699
Epoch 00050: val_accuracy did not improve from 0.90411
六、评估模型
1. Accuracy与Loss图
acc = train_model.history['accuracy']
val_acc = train_model.history['val_accuracy']
loss = train_model.history['loss']
val_loss = train_model.history['val_loss']
epochs_range = range(len(acc))
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()
可以看出模型的波动还是蛮大的，这主要是由于数据较少导致了（1462张图片9个类别），在数据扩充后，情况会得到有效改善的。
2. 混淆矩阵
from sklearn.metrics import confusion_matrix
import seaborn as sns
import pandas as pd
# 定义一个绘制混淆矩阵图的函数
def plot_cm(labels, predictions):
# 生成混淆矩阵
conf_numpy = confusion_matrix(labels, predictions)
# 将矩阵转化为 DataFrame
conf_df = pd.DataFrame(conf_numpy, index=class_names ,columns=class_names)
plt.figure(figsize=(8,7))
sns.heatmap(conf_df, annot=True, fmt="d", cmap="BuPu")
plt.title('混淆矩阵',fontsize=15)
plt.ylabel('真实值',fontsize=14)
plt.xlabel('预测值',fontsize=14)
val_pre = []
val_label = []
for images, labels in val_ds:#这里可以取部分验证数据（.take(1)）生成混淆矩阵
for image, label in zip(images, labels):
# 需要给图片增加一个维度
img_array = tf.expand_dims(image, 0)
# 使用模型预测图片中的人物
prediction = model.predict(img_array)
val_pre.append(class_names[np.argmax(prediction)])
val_label.append(class_names[label])
plot_cm(val_label, val_pre)
3. 各项指标评估
from sklearn import metrics
def test_accuracy_report(model):
print(metrics.classification_report(val_label, val_pre, target_names=class_names))
score = model.evaluate(val_ds, verbose=0)
print('Loss function: %s, accuracy:' % score[0], score[1])
test_accuracy_report(model)
precision recall f1-score support
buoy 1.00 0.67 0.80 3
cruise ship 0.86 0.82 0.84 22
ferry boat 1.00 0.50 0.67 6
freight boat 0.00 0.00 0.00 2
gondola 0.91 1.00 0.96 32
inflatable boat 0.00 0.00 0.00 2
kayak 0.76 0.86 0.81 22
paper boat 1.00 0.33 0.50 3
sailboat 0.88 0.96 0.92 54
accuracy 0.87 146
macro avg 0.71 0.57 0.61 146
weighted avg 0.85 0.87 0.85 146
Loss function: 0.5606985688209534, accuracy: 0.8698630332946777
