5天学习计划 第7周：咖啡豆识别_jie
前期工作
本文为365天深度学习训练营内部限免文章
参考本文所写记录性文章，请在文章开头保留以下内容
- 本文为365天深度学习训练营 中的学习记录博客
- 参考文章：365天深度学习训练营-第7周：咖啡豆识别（训练营内部成员可读）
- 原作者：K同学啊|接辅导、项目定制
我的环境：
- 语言环境：Python3.6.5
- 编译器：jupyter lab
- 深度学习环境：TensorFlow2.4.1
⏲往期文章：
-
5天学习计划-第6周：好莱坞明星识别
-
5天学习计划-第5周：运动鞋品牌识别
-
难度：夯实基础
-
语言：Python3、TensorFlow2
-
时间：9月5-9月9日
要求：
- 自己搭建VGG-16网络框架
- 调用官方的VGG-16网络框架
拔高（可选）：
- 验证集准确率达到100%
- 使用PPT画出VGG-16算法框架图（发论文需要这项技能）
探索（难度有点大）
- 在不影响准确率的前提下轻量化模型
设置GPU
如果使用的是CPU可以忽略这步
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")
导入数据
from tensorflow import keras
from tensorflow.keras import layers,models
import numpy as np
import matplotlib.pyplot as plt
import os,PIL,pathlib
# 这里需要更换成相应的地址
data_dir = "./49-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
import torch
import torch.nn as nn
import os,PIL,pathlib
from PIL import Image
from torchvision import transforms, datasets
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
数据预处理
加载数据
使用image_dataset_from_directory方法将磁盘中的数据加载到tf.data.Dataset中
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)
Found 1200 files belonging to 4 classes.
Using 960 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
"""
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)
Found 1200 files belonging to 4 classes.
Using 240 files for validation.
我们可以通过class_names输出数据集的标签。标签将按字母顺序对应于目录名称。
class_names = train_ds.class_names
print(class_names)
['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
- shuffle() ：打乱数据，关于此函数的详细介绍可以参考：https://zhuanlan.zhihu.com/p/42417456
- 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
train_ds = train_ds.cache().shuffle(1000).prefetch(buffer_size=AUTOTUNE)
val_ds = val_ds.cache().prefetch(buffer_size=AUTOTUNE)
normalization_layer = layers.experimental.preprocessing.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
在官方模型与自建模型之间进行二选一就可以了，选着一个注释掉另外一个。
VGG优缺点分析：
- VGG优点
VGG的结构非常简洁，整个网络都使用了同样大小的卷积核尺寸（3x3）和最大池化尺寸（2x2）。
- VGG缺点
1)训练时间过长，调参难度大。2)需要的存储容量大，不利于部署。例如存储VGG-16权重值文件的大小为500多MB，不利于安装到嵌入式系统中。
官方模型
官网模型调用这块我放到后面几篇文章中，下面主要讲一下VGG-16
from tensorflow.keras.models import Model
from tensorflow.keras.applications import VGG16
from tensorflow.keras.layers import Flatten, Dense
NUM_CLASSES = len(class_names)
model = tf.keras.applications.VGG16(include_top=False,weights='imagenet',input_shape=(224, 224, 3))
x = model.output
# 分类
x = Flatten(name='flatten')(x)
x = Dense(512, activation='relu', name='fc1-')(x)
x = Dense(128, activation='relu', name='fc2-')(x)
x = Dense(NUM_CLASSES, activation='softmax', name='predictions')(x)
model = Model(model.input, x, name="My-VGG16")
model.summary()
Model: "My-VGG16"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
input_1 (InputLayer) [(None, 224, 224, 3)] 0
block1_conv1 (Conv2D) (None, 224, 224, 64) 1792
block1_conv2 (Conv2D) (None, 224, 224, 64) 36928
block1_pool (MaxPooling2D) (None, 112, 112, 64) 0
block2_conv1 (Conv2D) (None, 112, 112, 128) 73856
block2_conv2 (Conv2D) (None, 112, 112, 128) 147584
block2_pool (MaxPooling2D) (None, 56, 56, 128) 0
block3_conv1 (Conv2D) (None, 56, 56, 256) 295168
block3_conv2 (Conv2D) (None, 56, 56, 256) 590080
block3_conv3 (Conv2D) (None, 56, 56, 256) 590080
block3_pool (MaxPooling2D) (None, 28, 28, 256) 0
block4_conv1 (Conv2D) (None, 28, 28, 512) 1180160
block4_conv2 (Conv2D) (None, 28, 28, 512) 2359808
block4_conv3 (Conv2D) (None, 28, 28, 512) 2359808
block4_pool (MaxPooling2D) (None, 14, 14, 512) 0
block5_conv1 (Conv2D) (None, 14, 14, 512) 2359808
block5_conv2 (Conv2D) (None, 14, 14, 512) 2359808
block5_conv3 (Conv2D) (None, 14, 14, 512) 2359808
block5_pool (MaxPooling2D) (None, 7, 7, 512) 0
flatten (Flatten) (None, 25088) 0
fc1- (Dense) (None, 512) 12845568
fc2- (Dense) (None, 128) 65664
predictions (Dense) (None, 4) 516
=================================================================
Total params: 27,626,436
Trainable params: 27,626,436
Non-trainable params: 0
_________________________________________________________________
自建模型
# 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()
2022-09-09 18:18:38.287389: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 411041792 exceeds 10% of free system memory.
Model: "model"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
input_1 (InputLayer) [(None, 224, 224, 3)] 0
block1_conv1 (Conv2D) (None, 224, 224, 64) 1792
block1_conv2 (Conv2D) (None, 224, 224, 64) 36928
block1_pool (MaxPooling2D) (None, 112, 112, 64) 0
block2_conv1 (Conv2D) (None, 112, 112, 128) 73856
block2_conv2 (Conv2D) (None, 112, 112, 128) 147584
block2_pool (MaxPooling2D) (None, 56, 56, 128) 0
block3_conv1 (Conv2D) (None, 56, 56, 256) 295168
block3_conv2 (Conv2D) (None, 56, 56, 256) 590080
block3_conv3 (Conv2D) (None, 56, 56, 256) 590080
block3_pool (MaxPooling2D) (None, 28, 28, 256) 0
block4_conv1 (Conv2D) (None, 28, 28, 512) 1180160
block4_conv2 (Conv2D) (None, 28, 28, 512) 2359808
block4_conv3 (Conv2D) (None, 28, 28, 512) 2359808
block4_pool (MaxPooling2D) (None, 14, 14, 512) 0
block5_conv1 (Conv2D) (None, 14, 14, 512) 2359808
block5_conv2 (Conv2D) (None, 14, 14, 512) 2359808
block5_conv3 (Conv2D) (None, 14, 14, 512) 2359808
block5_pool (MaxPooling2D) (None, 7, 7, 512) 0
flatten (Flatten) (None, 25088) 0
fc1 (Dense) (None, 4096) 102764544
fc2 (Dense) (None, 4096) 16781312
predictions (Dense) (None, 4) 16388
=================================================================
Total params: 134,276,932
Trainable params: 134,276,932
Non-trainable params: 0
_________________________________________________________________
2022-09-09 18:18:38.623610: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 411041792 exceeds 10% of free system memory.
2022-09-09 18:18:38.650108: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 411041792 exceeds 10% of free system memory.
网络结构图
参加了365天深度学习训练营的同学可以在语雀中查看网络结构图
VGG-16的结构说明：
● 13个卷积层（Convolutional Layer），分别用blockX_convX表示
● 3个全连接层（Fully connected Layer），分别用fcX与predictions表示
● 5个池化层（Pool layer），分别用blockX_pool表示
VGG-16包含了16个隐藏层（13个卷积层和3个全连接层），故称为VGG-16
自己画的结构图
编译
设置初始学习率
# 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=0.99, # 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'])
设置早停
model.compile(optimizer=opt,
loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
metrics=['accuracy'])
from tensorflow.keras.callbacks import ModelCheckpoint,EarlyStopping
checkpointer = ModelCheckpoint('best_model2.h5',
monitor='val_accuracy',
verbose=1,
save_best_only=True,
save_weights_only=True)
earlystopper = EarlyStopping(monitor='val_accuracy',
min_delta=0.001,
patience=10,
verbose=1 )
训练模型
epochs = 14
history = model.fit(train_ds,validation_data=val_ds,
epochs=epochs,callbacks=[checkpointer,earlystopper])
Epoch 1/14
/home/liangjie/anaconda3/lib/python3.9/site-packages/keras/backend.py:5581: 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(
2022-09-10 07:06:04.530289: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 411041792 exceeds 10% of free system memory.
2022-09-10 07:06:04.677350: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 411041792 exceeds 10% of free system memory.
2022-09-10 07:06:05.129971: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 205520896 exceeds 10% of free system memory.
2022-09-10 07:06:05.163211: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 205520896 exceeds 10% of free system memory.
2022-09-10 07:06:10.731038: W tensorflow/core/framework/cpu_allocator_impl.cc:82] Allocation of 205520896 exceeds 10% of free system memory.
30/30 [==============================] - ETA: 0s - loss: 0.7320 - accuracy: 0.6677
Epoch 1: val_accuracy improved from -inf to 0.91667, saving model to best_model2.h5
30/30 [==============================] - 267s 9s/step - loss: 0.7320 - accuracy: 0.6677 - val_loss: 0.2218 - val_accuracy: 0.9167
Epoch 2/14
30/30 [==============================] - ETA: 0s - loss: 0.1085 - accuracy: 0.9635
Epoch 2: val_accuracy improved from 0.91667 to 0.95833, saving model to best_model2.h5
30/30 [==============================] - 262s 9s/step - loss: 0.1085 - accuracy: 0.9635 - val_loss: 0.1196 - val_accuracy: 0.9583
Epoch 3/14
30/30 [==============================] - ETA: 0s - loss: 0.1225 - accuracy: 0.9552
Epoch 3: val_accuracy did not improve from 0.95833
30/30 [==============================] - 261s 9s/step - loss: 0.1225 - accuracy: 0.9552 - val_loss: 0.3096 - val_accuracy: 0.8917
Epoch 4/14
30/30 [==============================] - ETA: 0s - loss: 0.0507 - accuracy: 0.9823
Epoch 4: val_accuracy improved from 0.95833 to 0.99167, saving model to best_model2.h5
30/30 [==============================] - 261s 9s/step - loss: 0.0507 - accuracy: 0.9823 - val_loss: 0.0328 - val_accuracy: 0.9917
Epoch 5/14
30/30 [==============================] - ETA: 0s - loss: 0.1811 - accuracy: 0.9396
Epoch 5: val_accuracy did not improve from 0.99167
30/30 [==============================] - 260s 9s/step - loss: 0.1811 - accuracy: 0.9396 - val_loss: 0.1274 - val_accuracy: 0.9542
Epoch 6/14
30/30 [==============================] - ETA: 0s - loss: 0.0752 - accuracy: 0.9760
Epoch 6: val_accuracy did not improve from 0.99167
30/30 [==============================] - 261s 9s/step - loss: 0.0752 - accuracy: 0.9760 - val_loss: 0.1186 - val_accuracy: 0.9792
Epoch 7/14
30/30 [==============================] - ETA: 0s - loss: 0.0767 - accuracy: 0.9677
Epoch 7: val_accuracy did not improve from 0.99167
30/30 [==============================] - 259s 9s/step - loss: 0.0767 - accuracy: 0.9677 - val_loss: 0.0747 - val_accuracy: 0.9708
Epoch 8/14
30/30 [==============================] - ETA: 0s - loss: 0.0149 - accuracy: 0.9969
Epoch 8: val_accuracy did not improve from 0.99167
30/30 [==============================] - 262s 9s/step - loss: 0.0149 - accuracy: 0.9969 - val_loss: 0.0277 - val_accuracy: 0.9875
Epoch 9/14
30/30 [==============================] - ETA: 0s - loss: 0.0020 - accuracy: 1.0000
Epoch 9: val_accuracy improved from 0.99167 to 1.00000, saving model to best_model2.h5
30/30 [==============================] - 263s 9s/step - loss: 0.0020 - accuracy: 1.0000 - val_loss: 0.0027 - val_accuracy: 1.0000
Epoch 10/14
30/30 [==============================] - ETA: 0s - loss: 2.2677e-04 - accuracy: 1.0000
Epoch 10: val_accuracy did not improve from 1.00000
30/30 [==============================] - 259s 9s/step - loss: 2.2677e-04 - accuracy: 1.0000 - val_loss: 0.0018 - val_accuracy: 1.0000
Epoch 11/14
30/30 [==============================] - ETA: 0s - loss: 1.1921e-04 - accuracy: 1.0000
Epoch 11: val_accuracy did not improve from 1.00000
30/30 [==============================] - 259s 9s/step - loss: 1.1921e-04 - accuracy: 1.0000 - val_loss: 0.0014 - val_accuracy: 1.0000
Epoch 12/14
30/30 [==============================] - ETA: 0s - loss: 7.9640e-05 - accuracy: 1.0000
Epoch 12: val_accuracy did not improve from 1.00000
30/30 [==============================] - 258s 9s/step - loss: 7.9640e-05 - accuracy: 1.0000 - val_loss: 9.7935e-04 - val_accuracy: 1.0000
Epoch 13/14
30/30 [==============================] - ETA: 0s - loss: 5.4854e-05 - accuracy: 1.0000
Epoch 13: val_accuracy did not improve from 1.00000
30/30 [==============================] - 259s 9s/step - loss: 5.4854e-05 - accuracy: 1.0000 - val_loss: 6.7156e-04 - val_accuracy: 1.0000
Epoch 14/14
30/30 [==============================] - ETA: 0s - loss: 3.8973e-05 - accuracy: 1.0000
Epoch 14: val_accuracy did not improve from 1.00000
30/30 [==============================] - 258s 9s/step - loss: 3.8973e-05 - accuracy: 1.0000 - val_loss: 5.2667e-04 - val_accuracy: 1.0000
可视化结果
from pyecharts.charts import *
import pyecharts.options as opts
from pyecharts.globals import ThemeType
loss = history.history['loss']
val_loss = history.history['val_loss']
acc = history.history['accuracy']
val_acc = history.history['val_accuracy']
line_loss = Line()
line_loss.add_xaxis([i for i in range(20)])
line_loss.add_yaxis('loss', loss, label_opts=opts.LabelOpts(is_show=False))
line_loss.add_yaxis('val_loss', val_loss, label_opts=opts.LabelOpts(is_show=False))
line_loss.set_global_opts(legend_opts=opts.LegendOpts(pos_top='5%',pos_left='20%'),
tooltip_opts=opts.TooltipOpts(trigger="axis", axis_pointer_type="line"))
line_acc = Line()
line_acc.add_xaxis([i for i in range(20)])
line_acc.add_yaxis('accuracy', acc, label_opts=opts.LabelOpts(is_show=False))
line_acc.add_yaxis('val_accuracy', val_acc, label_opts=opts.LabelOpts(is_show=False))
line_acc.set_global_opts(title_opts=opts.TitleOpts('模型训练过程效果记录', pos_left='center'),
legend_opts=opts.LegendOpts(pos_top='5%', pos_left='65%'),
yaxis_opts=opts.AxisOpts(is_scale=True),
tooltip_opts=opts.TooltipOpts(trigger="axis", axis_pointer_type="line"))
grid = Grid(init_opts=opts.InitOpts(theme=ThemeType.CHALK))
grid.add(line_loss,grid_opts=opts.GridOpts(pos_left='5%', pos_right='55%'))
grid.add(line_acc,grid_opts=opts.GridOpts(pos_left='55%', pos_right='5%'))
grid.render_notebook()
# 模型的评估
acc = history.history['accuracy']
val_acc = history.history['val_accuracy']
loss = history.history['loss']
val_loss = history.history['val_loss']
epochs_range = range(len(loss))
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()
总结
源代码基础上，修改了哪些
设置了早停，epochs从20减少到14；利用VGG官方模型
有哪些改进
源代码结果：
改进后的效果：
修改结果前：val_accuracy: 0.96
修改结果后：val_accuracy: 1.00