第T7周：咖啡豆识别

于 2024-11-22 09:10:38 发布
阅读量670 收藏 14
点赞数 8
文章标签： 机器学习 python
版权声明：本文为博主原创文章，遵循 CC 4.0 BY-SA 版权协议，转载请附上原文出处链接和本声明。
本文链接：https://blog.youkuaiyun.com/2405_87523719/article/details/143960972
版权

🚀我的环境：

  • 语言环境：python 3.12.6
  • 编译器：jupyter lab
  • 深度学习环境：TensorFlow 2.17.0

前期准备

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 = "d:/Users/yxy/Desktop/T7data"
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

import tensorflow as tf

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.youkuaiyun.com/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.

class_names = train_ds.class_names
print(class_names)

['Dark', 'Green', 'Light', 'Medium']

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")

在这里插入图片描述

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)

from tensorflow.keras import layers

# 定义归一化层
normalization_layer = layers.Rescaling(1./255)

# 在数据集中应用归一化
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

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()

Model: "functional"

┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓
┃ Layer (type)                         ┃ Output Shape                ┃         Param # ┃
┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩
│ input_layer (InputLayer)             │ (None, 224, 224, 3)         │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block1_conv1 (Conv2D)                │ (None, 224, 224, 64)        │           1,792 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block1_conv2 (Conv2D)                │ (None, 224, 224, 64)        │          36,928 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block1_pool (MaxPooling2D)           │ (None, 112, 112, 64)        │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block2_conv1 (Conv2D)                │ (None, 112, 112, 128)       │          73,856 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block2_conv2 (Conv2D)                │ (None, 112, 112, 128)       │         147,584 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block2_pool (MaxPooling2D)           │ (None, 56, 56, 128)         │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block3_conv1 (Conv2D)                │ (None, 56, 56, 256)         │         295,168 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block3_conv2 (Conv2D)                │ (None, 56, 56, 256)         │         590,080 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block3_conv3 (Conv2D)                │ (None, 56, 56, 256)         │         590,080 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block3_pool (MaxPooling2D)           │ (None, 28, 28, 256)         │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block4_conv1 (Conv2D)                │ (None, 28, 28, 512)         │       1,180,160 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block4_conv2 (Conv2D)                │ (None, 28, 28, 512)         │       2,359,808 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block4_conv3 (Conv2D)                │ (None, 28, 28, 512)         │       2,359,808 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block4_pool (MaxPooling2D)           │ (None, 14, 14, 512)         │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block5_conv1 (Conv2D)                │ (None, 14, 14, 512)         │       2,359,808 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block5_conv2 (Conv2D)                │ (None, 14, 14, 512)         │       2,359,808 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block5_conv3 (Conv2D)                │ (None, 14, 14, 512)         │       2,359,808 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ block5_pool (MaxPooling2D)           │ (None, 7, 7, 512)           │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ flatten (Flatten)                    │ (None, 25088)               │               0 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ fc1 (Dense)                          │ (None, 4096)                │     102,764,544 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ fc2 (Dense)                          │ (None, 4096)                │      16,781,312 │
├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤
│ predictions (Dense)                  │ (None, 4)                   │          16,388 │
└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘

 Total params: 134,276,932 (512.23 MB)

 Trainable params: 134,276,932 (512.23 MB)

 Non-trainable params: 0 (0.00 B)

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=False),
              metrics=['accuracy'])

训练模型

epochs = 20

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

Epoch 1/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m434s[0m 14s/step - accuracy: 0.2596 - loss: 1.3600 - val_accuracy: 0.5708 - val_loss: 1.1062
Epoch 2/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m420s[0m 14s/step - accuracy: 0.5386 - loss: 0.8870 - val_accuracy: 0.7667 - val_loss: 0.5559
Epoch 3/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m422s[0m 14s/step - accuracy: 0.7285 - loss: 0.5818 - val_accuracy: 0.8250 - val_loss: 0.4575
Epoch 4/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m419s[0m 14s/step - accuracy: 0.8279 - loss: 0.4456 - val_accuracy: 0.8667 - val_loss: 0.4078
Epoch 5/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m417s[0m 14s/step - accuracy: 0.8823 - loss: 0.3405 - val_accuracy: 0.9500 - val_loss: 0.1400
Epoch 6/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m417s[0m 14s/step - accuracy: 0.9214 - loss: 0.1954 - val_accuracy: 0.9417 - val_loss: 0.1281
Epoch 7/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m418s[0m 14s/step - accuracy: 0.9480 - loss: 0.1663 - val_accuracy: 0.9542 - val_loss: 0.1312
Epoch 8/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m425s[0m 14s/step - accuracy: 0.9314 - loss: 0.1968 - val_accuracy: 0.8458 - val_loss: 0.4145
Epoch 9/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m415s[0m 14s/step - accuracy: 0.9265 - loss: 0.1828 - val_accuracy: 0.9833 - val_loss: 0.0823
Epoch 10/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m415s[0m 14s/step - accuracy: 0.9782 - loss: 0.0610 - val_accuracy: 0.9792 - val_loss: 0.0655
Epoch 11/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m415s[0m 14s/step - accuracy: 0.9709 - loss: 0.0904 - val_accuracy: 0.9542 - val_loss: 0.1089
Epoch 12/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m418s[0m 14s/step - accuracy: 0.9720 - loss: 0.0836 - val_accuracy: 0.9583 - val_loss: 0.1337
Epoch 13/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m420s[0m 14s/step - accuracy: 0.9893 - loss: 0.0385 - val_accuracy: 0.9500 - val_loss: 0.1916
Epoch 14/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m420s[0m 14s/step - accuracy: 0.9572 - loss: 0.0980 - val_accuracy: 0.9833 - val_loss: 0.0441
Epoch 15/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m415s[0m 14s/step - accuracy: 0.9942 - loss: 0.0228 - val_accuracy: 0.9292 - val_loss: 0.2179
Epoch 16/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m413s[0m 14s/step - accuracy: 0.9403 - loss: 0.2382 - val_accuracy: 0.9625 - val_loss: 0.0909
Epoch 17/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m416s[0m 14s/step - accuracy: 0.9865 - loss: 0.0377 - val_accuracy: 0.9708 - val_loss: 0.0673
Epoch 18/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m418s[0m 14s/step - accuracy: 0.9912 - loss: 0.0269 - val_accuracy: 0.9625 - val_loss: 0.0622
Epoch 19/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m425s[0m 14s/step - accuracy: 0.9913 - loss: 0.0235 - val_accuracy: 0.9667 - val_loss: 0.1793
Epoch 20/20
[1m30/30[0m [32m━━━━━━━━━━━━━━━━━━━━[0m[37m[0m [1m426s[0m 14s/step - accuracy: 0.9609 - loss: 0.1167 - val_accuracy: 0.9750 - val_loss: 0.0928

可视化结果

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()

在这里插入图片描述
总结
1.shuffle() ：打乱数据，
prefetch() ：预取数据，加速运行，
cache() ：将数据集缓存到内存当中，加速运行
2.VGG优点：
结构简单：使用连续的小卷积核（3×3）堆叠，便于理解和实现。
迁移学习效果好：预训练模型广泛用于其他任务，效果优秀。
特征表达强：深层网络捕获高质量特征，对视觉任务表现出色。
VGG缺点：
参数量大：权重较多，训练和存储成本高。
计算量大：推理速度慢，对硬件要求较高。
梯度消失风险：网络较深，容易出现梯度消失问题。

评论
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付：10.00元
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付元
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明：

1.余额是钱包充值的虚拟货币，按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载，可以购买VIP、付费专栏及课程。

余额充值