第J6周：RenseNeXt-50实战

• 🍨 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
  

电脑环境：
语言环境：Python 3.8.0
深度学习环境：tensorflow 2.17.0

一、前言

本次使用的数据集是猴痘病毒数据集。

1、结构改进

上图是ResNet（左）与ResNeXt（右）block的差异。在ResNet中，输入的具有256个通道的特征经过1×1卷积压缩4倍到64个通道，之后3×3的卷积核用于处理特征，经1×1卷积扩大通道数与原特征残差连接后输出。ResNeXt也是相同的处理策略，但在ResNeXt中，输入的具有256个通道的特征被分为32个组，每组被压缩64倍到4个通道后进行处理。32个组相加后与原特征残差连接后输出。这里cardinatity指的是一个block中所具有的相同分支的数目。

2、分组卷积

ResNeXt中采用的分组卷积简单来说就是将特征图分为不同的组，再对每组特征图分别进行卷积，这个操作可以有效的降低计算量。
在分组卷积中，每个卷积核只处理部分通道，比如下图中，红色卷积核只处理红色的通道，绿色卷积核只处理绿色通道，黄色卷积核只处理黄色通道。此时每个卷积核有2个通道，每个卷积核生成一张特征图。

二、前期工作

1.设置GPU

from tensorflow import keras
from keras import layers, models
import os, PIL, pathlib
import matplotlib.pyplot as plt
import tensorflow as tfgpus = tf.config.list_physical_devices("GPU")if gpus:gpu0 = gpus[0] #如果有多个GPU，仅使用第0个GPUtf.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")

2. 导入数据

data_dir = "./data/"
data_dir = pathlib.Path(data_dir)

3. 查看数据

image_count = len(list(data_dir.glob('*/*.jpg')))
print("图片总数为：",image_count)

输出：图片总数为： 2142

三、数据预处理

1、加载数据

batch_size = 8
img_height = 224
img_width = 224train_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)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

2、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)

四、构建网络

1、导入包

import numpy as np
from tensorflow.keras.preprocessing.image import ImageDataGenerator
from tensorflow.keras.utils import to_categorical
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Input, Dense, Dropout, Conv2D, MaxPool2D, Flatten, GlobalAvgPool2D, concatenate, \
BatchNormalization, Activation, Add, ZeroPadding2D, Lambda
from tensorflow.keras.optimizers import Adam
from tensorflow.keras.layers import ReLU
import matplotlib.pyplot as plt
from tensorflow.keras.callbacks import LearningRateScheduler
from tensorflow.keras.models import Model

2、分组卷积模块

# 定义分组卷积
def grouped_convolution_block(init_x, strides, groups, g_channels):group_list = []# 分组进行卷积for c in range(groups):# 分组取出数据x = Lambda(lambda x: x[:, :, :, c * g_channels:(c + 1) * g_channels])(init_x)# 分组进行卷积x = Conv2D(filters=g_channels, kernel_size=(3, 3),strides=strides, padding='same', use_bias=False)(x)# 存入listgroup_list.append(x)# 合并list中的数据group_merage = concatenate(group_list, axis=3)x = BatchNormalization(epsilon=1.001e-5)(group_merage)x = ReLU()(x)return x

3、残差单元

# 定义残差单元
def block(x, filters, strides=1, groups=32, conv_shortcut=True):if conv_shortcut:shortcut = Conv2D(filters * 2, kernel_size=(1, 1), strides=strides, padding='same', use_bias=False)(x)# epsilon为BN公式中防止分母为零的值shortcut = BatchNormalization(epsilon=1.001e-5)(shortcut)else:# identity_shortcutshortcut = x# 三层卷积层x = Conv2D(filters=filters, kernel_size=(1, 1), strides=1, padding='same', use_bias=False)(x)x = BatchNormalization(epsilon=1.001e-5)(x)x = ReLU()(x)# 计算每组的通道数g_channels = int(filters / groups)# 进行分组卷积x = grouped_convolution_block(x, strides, groups, g_channels)x = Conv2D(filters=filters * 2, kernel_size=(1, 1), strides=1, padding='same', use_bias=False)(x)x = BatchNormalization(epsilon=1.001e-5)(x)x = Add()([x, shortcut])x = ReLU()(x)return x

4、堆叠残差单元

# 堆叠残差单元
def stack(x, filters, blocks, strides, groups=32):# 每个stack的第一个block的残差连接都需要使用1*1卷积升维x = block(x, filters, strides=strides, groups=groups)for i in range(blocks):x = block(x, filters, groups=groups, conv_shortcut=False)return x

5、搭建ResNeXt-50网络

# 定义ResNext50(32*4d)网络
def ResNext50(input_shape, num_classes):inputs = Input(shape=input_shape)# 填充3圈0，[224,224,3]->[230,230,3]x = ZeroPadding2D((3, 3))(inputs)x = Conv2D(filters=64, kernel_size=(7, 7), strides=2, padding='valid')(x)x = BatchNormalization(epsilon=1.001e-5)(x)x = ReLU()(x)# 填充1圈0x = ZeroPadding2D((1, 1))(x)x = MaxPool2D(pool_size=(3, 3), strides=2, padding='valid')(x)# 堆叠残差结构x = stack(x, filters=128, blocks=2, strides=1)x = stack(x, filters=256, blocks=3, strides=2)x = stack(x, filters=512, blocks=5, strides=2)x = stack(x, filters=1024, blocks=2, strides=2)# 根据特征图大小进行全局平均池化x = GlobalAvgPool2D()(x)x = Dense(num_classes, activation='softmax')(x)# 定义模型model = Model(inputs=inputs, outputs=x)return model

6、查看模型摘要

model=ResNext50(input_shape=(224,224,3),num_classes=1000)
model.summary()

五、编译

# 设置优化器
opt = tf.keras.optimizers.Adam(learning_rate=1e-4)
model.compile(optimizer=opt,loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),metrics=['accuracy'])

六、训练模型

epochs = 20history = model.fit(train_ds,validation_data=val_ds,epochs=epochs)

Epoch 1/20
215/215 ━━━━━━━━━━━━━━━━━━━━ 292s 557ms/step - accuracy: 0.4838 - loss: 1.8304 - val_accuracy: 0.5701 - val_loss: 0.7159
..................................................................................
Epoch 18/20
215/215 ━━━━━━━━━━━━━━━━━━━━ 37s 173ms/step - accuracy: 0.9794 - loss: 0.0574 - val_accuracy: 0.8014 - val_loss: 0.6634
Epoch 19/20
215/215 ━━━━━━━━━━━━━━━━━━━━ 37s 173ms/step - accuracy: 0.9749 - loss: 0.0660 - val_accuracy: 0.7640 - val_loss: 0.7989
Epoch 20/20
215/215 ━━━━━━━━━━━━━━━━━━━━ 41s 175ms/step - accuracy: 0.9610 - loss: 0.1093 - val_accuracy: 0.7780 - val_loss: 0.6207

七、模型评估

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()