【深度学习TensorFlow（九）】剪刀石头布手势识别（多分类
2021-02-20 本文已影响0人
Geekero
学习自中国大学MOOC TensorFlow学习课程
一、加载数据
下载
!wget --no-check-certificate \
https://storage.googleapis.com/laurencemoroney-blog.appspot.com/rps.zip \
-O /tmp/rps.zip
!wget --no-check-certificate \
https://storage.googleapis.com/laurencemoroney-blog.appspot.com/rps-test-set.zip \
-O /tmp/rps-test-set.zip
'wget' 不是内部或外部命令，也不是可运行的程序
或批处理文件。
'wget' 不是内部或外部命令，也不是可运行的程序
或批处理文件。
解压
import os
import zipfile
local_zip = 'rps.zip'
zip_ref = zipfile.ZipFile(local_zip, 'r')
zip_ref.extractall('./')
zip_ref.close()
local_zip = 'rps-test-set.zip'
zip_ref = zipfile.ZipFile(local_zip, 'r')
zip_ref.extractall('./')
zip_ref.close()
查看数据结构
rock_dir = os.path.join('./rps/rock')
paper_dir = os.path.join('./rps/paper')
scissors_dir = os.path.join('./rps/scissors')
print('total training rock images:', len(os.listdir(rock_dir)))
print('total training paper images:', len(os.listdir(paper_dir)))
print('total training scissors images:', len(os.listdir(scissors_dir)))
rock_files = os.listdir(rock_dir)
print(rock_files[:10])
paper_files = os.listdir(paper_dir)
print(paper_files[:10])
scissors_files = os.listdir(scissors_dir)
print(scissors_files[:10])
total training rock images: 840
total training paper images: 840
total training scissors images: 840
['rock01-000.png', 'rock01-001.png', 'rock01-002.png', 'rock01-003.png', 'rock01-004.png', 'rock01-005.png', 'rock01-006.png', 'rock01-007.png', 'rock01-008.png', 'rock01-009.png']
['paper01-000.png', 'paper01-001.png', 'paper01-002.png', 'paper01-003.png', 'paper01-004.png', 'paper01-005.png', 'paper01-006.png', 'paper01-007.png', 'paper01-008.png', 'paper01-009.png']
['scissors01-000.png', 'scissors01-001.png', 'scissors01-002.png', 'scissors01-003.png', 'scissors01-004.png', 'scissors01-005.png', 'scissors01-006.png', 'scissors01-007.png', 'scissors01-008.png', 'scissors01-009.png']
可视化
%matplotlib inline
import matplotlib.pyplot as plt
import matplotlib.image as mpimg
pic_index = 2
next_rock = [os.path.join(rock_dir, fname)
for fname in rock_files[pic_index-2:pic_index]]
next_paper = [os.path.join(paper_dir, fname)
for fname in paper_files[pic_index-2:pic_index]]
next_scissors = [os.path.join(scissors_dir, fname)
for fname in scissors_files[pic_index-2:pic_index]]
for i, img_path in enumerate(next_rock+next_paper+next_scissors):
#print(img_path)
img = mpimg.imread(img_path)
plt.imshow(img)
plt.axis('Off')
plt.show()
output_3_0.png
output_3_1.png
output_3_2.png
output_3_3.png
output_3_4.png
output_3_5.png
二、构建神经网络并训练网络
import tensorflow as tf
import keras_preprocessing
from keras_preprocessing import image
from keras_preprocessing.image import ImageDataGenerator
TRAINING_DIR = "./rps/"
training_datagen = ImageDataGenerator(
rescale = 1./255,
rotation_range=40,
width_shift_range=0.2,
height_shift_range=0.2,
shear_range=0.2,
zoom_range=0.2,
horizontal_flip=True,
fill_mode='nearest')
VALIDATION_DIR = "./rps-test-set/"
validation_datagen = ImageDataGenerator(rescale = 1./255)
train_generator = training_datagen.flow_from_directory(
TRAINING_DIR,
target_size=(150,150),
class_mode='categorical'
)
validation_generator = validation_datagen.flow_from_directory(
VALIDATION_DIR,
target_size=(150,150),
class_mode='categorical'
)
model = tf.keras.models.Sequential([
# Note the input shape is the desired size of the image 150x150 with 3 bytes color
# This is the first convolution
tf.keras.layers.Conv2D(64, (3,3), activation='relu', input_shape=(150, 150, 3)),
tf.keras.layers.MaxPooling2D(2, 2),
# The second convolution
tf.keras.layers.Conv2D(64, (3,3), activation='relu'),
tf.keras.layers.MaxPooling2D(2,2),
# The third convolution
tf.keras.layers.Conv2D(128, (3,3), activation='relu'),
tf.keras.layers.MaxPooling2D(2,2),
# The fourth convolution
tf.keras.layers.Conv2D(128, (3,3), activation='relu'),
tf.keras.layers.MaxPooling2D(2,2),
# Flatten the results to feed into a DNN
tf.keras.layers.Flatten(),
tf.keras.layers.Dropout(0.5),
# 512 neuron hidden layer
tf.keras.layers.Dense(512, activation='relu'),
tf.keras.layers.Dense(3, activation='softmax') #三个输出，所有输出合起来等于1，每个分类的概率在0~1之间
])
model.summary()
model.compile(loss = 'categorical_crossentropy', optimizer='rmsprop', metrics=['accuracy']) #优化学习步长的算法rmsprop
history = model.fit_generator(train_generator, epochs=5, validation_data = validation_generator, verbose = 1)
model.save("rps.h5")
Found 2520 images belonging to 3 classes.
Found 372 images belonging to 3 classes.
Model: "sequential"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
conv2d (Conv2D) (None, 148, 148, 64) 1792
_________________________________________________________________
max_pooling2d (MaxPooling2D) (None, 74, 74, 64) 0
_________________________________________________________________
conv2d_1 (Conv2D) (None, 72, 72, 64) 36928
_________________________________________________________________
max_pooling2d_1 (MaxPooling2 (None, 36, 36, 64) 0
_________________________________________________________________
conv2d_2 (Conv2D) (None, 34, 34, 128) 73856
_________________________________________________________________
max_pooling2d_2 (MaxPooling2 (None, 17, 17, 128) 0
_________________________________________________________________
conv2d_3 (Conv2D) (None, 15, 15, 128) 147584
_________________________________________________________________
max_pooling2d_3 (MaxPooling2 (None, 7, 7, 128) 0
_________________________________________________________________
flatten (Flatten) (None, 6272) 0
_________________________________________________________________
dropout (Dropout) (None, 6272) 0
_________________________________________________________________
dense (Dense) (None, 512) 3211776
_________________________________________________________________
dense_1 (Dense) (None, 3) 1539
=================================================================
Total params: 3,473,475
Trainable params: 3,473,475
Non-trainable params: 0
_________________________________________________________________
D:\anaconda\envs\TF2_4\lib\site-packages\tensorflow\python\keras\engine\training.py:1844: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.
warnings.warn('`Model.fit_generator` is deprecated and '
Epoch 1/5
79/79 [==============================] - 53s 613ms/step - loss: 1.3453 - accuracy: 0.3601 - val_loss: 0.7386 - val_accuracy: 0.7151
Epoch 2/5
79/79 [==============================] - 26s 323ms/step - loss: 0.9369 - accuracy: 0.5563 - val_loss: 0.2506 - val_accuracy: 0.9704
Epoch 3/5
79/79 [==============================] - 27s 339ms/step - loss: 0.5901 - accuracy: 0.7418 - val_loss: 0.1306 - val_accuracy: 0.9892
Epoch 4/5
79/79 [==============================] - 26s 329ms/step - loss: 0.3498 - accuracy: 0.8712 - val_loss: 0.1069 - val_accuracy: 0.9731
Epoch 5/5
79/79 [==============================] - 26s 322ms/step - loss: 0.2504 - accuracy: 0.9016 - val_loss: 0.0632 - val_accuracy: 0.9677
评估模型性能
import matplotlib.pyplot as plt
acc = history.history['accuracy']
val_acc = history.history['val_accuracy']
loss = history.history['loss']
val_loss = history.history['val_accuracy']
epochs = range(len(acc))
plt.plot(epochs, acc, 'r', label='Training accuracy')
plt.plot(epochs, val_acc, 'b', label='Validation accuracy')
plt.title('Training and validation accuracy')
plt.legend(loc=0)
plt.figure()
plt.show()
<Figure size 432x288 with 0 Axes>
import numpy as np
from google.colab import files
from keras.preprocessing import image
uploaded = files.upload()
for fn in uploaded.keys():
# predicting images
path = fn
img = image.load_img(path, target_size=(150, 150))
x = image.img_to_array(img)
x = np.expand_dims(x, axis=0)
images = np.vstack([x])
classes = model.predict(images, batch_size=10)
print(fn)
print(classes)
import numpy as np
from keras.preprocessing import image
# predicting images
path = 'D:/Data/Tensorflow/第二部分/第八章图像多分类问题/rps-test-set/scissors/testscissors01-19.png'
img = image.load_img(path, target_size=(150, 150))
x = image.img_to_array(img)
x = np.expand_dims(x, axis=0)
images = np.vstack([x])
classes = model.predict(images, batch_size=10)
print(classes)
[[0. 0. 1.]]
img
output_8_0.png
#释放资源
import os, signal
#在Windows中，signal()只能叫SIGABRT， SIGFPE，SIGILL，SIGINT，SIGSEGV，或 SIGTERM。ValueError在其他情况下，将引发A。
#os.kill(os.getpid(), signal.SIGKILL)
os.kill(os.getpid(), signal.SIGINT)
