cats_vs_dogs kaggle - cnn 练习

Github: https://github.com/yjfiejd/CNN_learning/blob/master/cats_vs_dogs_jupyter_notebook.ipynb (jupyter notebook)

kaggle: https://www.kaggle.com/c/dogs-vs-cats

备注：由于mac电脑带不动，目前只训练了部分照片，这样快些出结果

思路：

#【1】#导入常用的安装包，确保安装好opencv, keras
#1）获得图片路径名称 + 标签 #路径里面不要有中文，伤不起啊！要养成习惯
#2）train中的dogs, cats
#3）#合成训练集数据
#4）#test中的数据
#5）#打乱训练集中的数据
#6）把所有的图片resize进64*64
#7）把图片数据变换为容易使用的numpy数组
#8）#成功的把图片路径输入，得到所有图片的已经resize后的array数组格式，和对应的label值


#【2】 CNN模型构造
#1） # 建造模型

#【3】 训练与预测
#1）## 每个epoch之后，存下loss，便于画出图
#2）# 跑模型
#3）# 画图

# -*- coding:utf8 -*-
# @TIME : 2018/5/6 上午7:46
# @Author : Allen
# @File : test-cats&dogs.py

#【1】#导入常用的安装包，确保安装好opencv, keras
import os, cv2, random
import numpy as np
import pandas as pd

import matplotlib.pyplot as plt
from matplotlib import ticker
import seaborn as sns

from keras.models import Sequential
from keras.layers import Input, Dropout, Flatten, Convolution2D, MaxPooling2D, Dense, Activation
from keras.optimizers import RMSprop
from keras.callbacks import ModelCheckpoint, Callback, EarlyStopping
from keras.utils import np_utils

#1）获得图片路径名称 + 标签
#路径里面不要有中文，伤不起啊！要养成习惯
TRAIN_DIR = '/Users/a1/Desktop/learning/kaggel_06/cat_dog/train_small/'
TEST_DIR = '/Users/a1/Desktop/learning/kaggel_06/cat_dog/test_samll/'

#train中的dogs, cats
train_dogs = [(TRAIN_DIR+i, 1) for i in os.listdir(TRAIN_DIR) if 'dog' in i]
train_cats = [(TRAIN_DIR+i, 0) for i in os.listdir(TRAIN_DIR) if 'cat' in i]

#合成训练集数据
train_images = train_dogs + train_cats

#test中的数据
test_images = [(TEST_DIR+i, -1) for i in os.listdir(TEST_DIR)]

#打乱训练集中的数据
random.shuffle(train_images)

#这里其实要学习如何分批次读取数据，保护内存不爆，一会再试试

#2) 把所有的图片resize进64*64
ROWS = 64
COLS = 64
def read_image(tuple_set):
    file_path = tuple_set[0]
    label = tuple_set[1]
    # print(file_path)
    # print(label)
    # Opencv图像的读取、显示和保存：https://blog.csdn.net/djcxym/article/details/52097812
    img = cv2.imread(file_path, cv2.IMREAD_COLOR)  # 读取图片 cv2.imread(filepath,flags)

    # img.all()
    return cv2.resize(img, (ROWS, COLS), interpolation=cv2.INTER_CUBIC), label
# 已解决错误：【python】error: (-215) ssize.width > 0 && ssize.height > 0 in function cv::resize：https://blog.csdn.net/qq_30214939/article/details/77432167
# 这个错误的原因是因为我刚没有把dogs+cats的数据合并起来，导致输入的图片都路径不对，导致imread()没有读取图片，返回None，我去！

#3) 把图片数据变换为容易使用的numpy数组
CHANNELS = 3
def prep_data(images):
    no_images = len(images)
    print(no_images) #长度为102
    #ndarray对象属性:https://blog.csdn.net/scorpion_zs/article/details/52526310
    data = np.ndarray((no_images, CHANNELS, ROWS, COLS), dtype=np.uint8)
    labels = []
    for i, image_file in enumerate(images):
        print('*********')
        print(i)
        print(image_file)
        print('#########')
        image, label = read_image(image_file)
        data[i] = image.T
        labels.append(label)
    return data, labels
#成功的把图片路径输入，得到所有图片的已经resize后的array数组格式，和对应的label值
x_train, y_train = prep_data(train_images)

*********
# 0
# ('/Users/a1/Desktop/learning/kaggel_06/cat_dog/train_small/cat.3746.jpg', 0)
# #########
# *********
# 1
# ('/Users/a1/Desktop/learning/kaggel_06/cat_dog/train_small/cat.3733.jpg', 0)
# #########
# *********
# 2
# ('/Users/a1/Desktop/learning/kaggel_06/cat_dog/train_small/cat.3717.jpg', 0)
# #########
# *********


#展示一下返回的data， label长啥样子
# [[[[111 111 111 ... 198 180 181]
#    [110 110 108 ... 187 186 169]
#    [109 109 108 ... 197 153 173]
#    ...
#    [160 159 161 ... 169 165 177]
#    [158 159 159 ... 165 173 187]
#    [157 159 158 ... 155 142 175]]

#   [[112 113 115 ... 202 184 189]
#    [113 112 109 ... 194 193 176]
#    [112 112 111 ... 201 157 179]
#    ...
#    [158 157 159 ... 180 176 186]
#    [156 157 157 ... 176 184 194]
#    [155 157 156 ... 166 153 183]]
#  这里省略了很多

#看下label结果
# [0, 0, 0, 0, 1, 1, 1, 0, 1, 0, 1, 1, 0, 1, 0, 0, 0, 0, 1, 0, 0, 1, 1, 0, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 1, 0, 1, 1, 1, 0, 0, 1, 0, 0, 1, 1, 0, 1, 0, 1, 0, 1, 0, 1, 1, 1, 1, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 0, 1, 1, 0, 0, 1, 1, 1, 0, 1, 0, 0, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 0, 1]

x_test, y_shit = prep_data(test_images)

print(x_train.shape)
print(x_test.shape)

##【2】 CNN模型构造

#这里有个报错提示:ValueError: Negative dimension size caused by subtracting 2 from 1 for 'MaxPool_1' (op: 'MaxPool') w，需要添加下面两行代码：https://blog.csdn.net/w5688414/article/details/78780440
from keras import backend as K
K.set_image_dim_ordering('th')

optimizer = RMSprop(lr=1e-4)
objective = 'binary_crossentropy'

# 建造模型
model = Sequential()

model.add(Convolution2D(32, 3, 3, border_mode='same', input_shape=(3, ROWS, COLS), activation='relu'))
model.add(Convolution2D(32, 3, 3, border_mode='same', activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))

model.add(Convolution2D(64, 3, 3, border_mode='same', activation='relu'))
model.add(Convolution2D(64, 3, 3, border_mode='same', activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))

model.add(Convolution2D(128, 3, 3, border_mode='same', activation='relu'))
model.add(Convolution2D(128, 3, 3, border_mode='same', activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))

model.add(Convolution2D(256, 3, 3, border_mode='same', activation='relu'))
model.add(Convolution2D(256, 3, 3, border_mode='same', activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))

model.add(Flatten())
model.add(Dense(256, activation='relu'))
model.add(Dropout(0.5))

model.add(Dense(256, activation='relu'))
model.add(Dropout(0.5))

model.add(Dense(1))
model.add(Activation('sigmoid'))

model.compile(loss=objective, optimizer=optimizer, metrics=['accuracy'])

# 【3】 训练与预测

nb_epoch = 10
batch_size = 10


## 每个epoch之后，存下loss，便于画出图
class LossHistory(Callback):
    def on_train_begin(self, logs={}):
        self.losses = []
        self.val_losses = []

    def on_epoch_end(self, batch, logs={}):
        self.losses.append(logs.get('loss'))
        self.val_losses.append(logs.get('val_loss'))


early_stopping = EarlyStopping(monitor='val_loss', patience=3, verbose=1, mode='auto')

# 跑模型
history = LossHistory()

model.fit(x_train, y_train, batch_size=batch_size, nb_epoch=nb_epoch,
          validation_split=0.2, verbose=0, shuffle=True, callbacks=[history, early_stopping])

predictions = model.predict(x_test, verbose=0)

loss = history.losses
val_loss = history.val_losses

plt.xlabel('Epochs')
plt.ylabel('Loss')
plt.title('VGG-16 Loss Trend')
plt.plot(loss, 'blue', label='Training Loss')
plt.plot(val_loss, 'green', label='Validation Loss')
plt.xticks(range(0,nb_epoch)[0::2])
plt.legend()
plt.show()