深度学习图像分类器结构配置

Question

你好，我是机器学习的新手，只知道一些基本知识和事情应该如何工作。因此，我看了这个关于Python、TensorFlow和Keras教程的深度学习教程，并获得了以下代码

import tensorflow as tf
from keras.datasets import cifar10
from keras.preprocessing.image import ImageDataGenerator
from keras.models import Sequential
from keras.layers import Dense, Dropout, Activation, Flatten
from keras.layers import Conv2D, MaxPooling2D
import numpy as np
import matplotlib.pyplot as plt
import os
import cv2

DATADIR = "C:/Users/Acer/imagerec/MRI"

CATEGORIES = ["yes", "no"]

for category in CATEGORIES:
    path = os.path.join(DATADIR,category)
    for img in os.listdir(path):
        img_array = cv2.imread(os.path.join(path,img) ,cv2.IMREAD_GRAYSCALE)
        plt.imshow(img_array, cmap='gray')


        break
    break
print(img_array)
print(img_array.shape)

IMG_SIZE = 224

new_array = cv2.resize(img_array, (IMG_SIZE, IMG_SIZE))
plt.imshow(new_array, cmap='gray')


training_data = []

def create_training_data():
    for category in CATEGORIES:
        path = os.path.join(DATADIR, category)
        class_num = CATEGORIES.index(category)
        for img in os.listdir(path):
            try:
                img_array = cv2.imread(os.path.join(path, img), cv2.IMREAD_GRAYSCALE)
                new_array = cv2.resize(img_array, (IMG_SIZE, IMG_SIZE))
                training_data.append([new_array, class_num])
            except Exception as e:
                pass


create_training_data()

print(len(training_data))

import random

random.shuffle(training_data)
for sample in training_data[:10]:
    print(sample[1])

X = []
y = []
for features, label in training_data:
    X.append(features)
    y.append(label)

X = np.array(X).reshape(-1, IMG_SIZE, IMG_SIZE, 1)

import pickle

pickle_in = open("X.pickle","rb")
X = pickle.load(pickle_in)



pickle_in = open("y.pickle","rb")
y = pickle.load(pickle_in)

X = X/255.0

model = Sequential()

model.add(Conv2D(256, (3, 3), input_shape=X.shape[1:]))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))

model.add(Conv2D(256, (3, 3)))
model.add(Activation('relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))

model.add(Flatten())

model.add(Dense(64))
model.add(Activation('relu'))

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

model.compile(loss='binary_crossentropy',
              optimizer='adam',
              metrics=['accuracy'])
model.fit(X, y, batch_size=15, epochs=20, validation_split=0.1)

from sklearn.metrics import confusion_matrix
pred = model.predict(X)
pred = np.round(pred)

conf = confusion_matrix(y, pred)

import seaborn as sns
sns.heatmap(conf, annot=True)

plt.show()

因此，运行这些代码给了我76.9%的验证准确率，我需要做的是将代码的模型转换为VGG16、VGG19和mobilenet，但是我不知道如何导入经过预先训练的模型，所以我决定建立自己的模型并对其进行培训，所以我查看了VGG16和VGG19的体系结构，我查看了多少conv和was池，并提出了这个代码。

import tensorflow as tf
from keras.datasets import cifar10
from keras.preprocessing.image import ImageDataGenerator
from keras.models import Sequential
from keras.layers import Dense, Dropout, Activation, Flatten
from keras.layers import Conv2D, MaxPooling2D
import numpy as np
import matplotlib.pyplot as plt
import os
import cv2

DATADIR = "C:/Users/Acer/imagerec/EDA"

CATEGORIES = ["yes", "no"]

for category in CATEGORIES:
    path = os.path.join(DATADIR,category)
    for img in os.listdir(path):
        img_array = cv2.imread(os.path.join(path,img) ,cv2.IMREAD_GRAYSCALE)
        plt.imshow(img_array, cmap='gray')
        plt.show()

        break
    break
print(img_array)
print(img_array.shape)

IMG_SIZE = 224

new_array = cv2.resize(img_array, (IMG_SIZE, IMG_SIZE))
plt.imshow(new_array, cmap='gray')
plt.show()

training_data = []

def create_training_data():
    for category in CATEGORIES:
        path = os.path.join(DATADIR, category)
        class_num = CATEGORIES.index(category)
        for img in os.listdir(path):
            try:
                img_array = cv2.imread(os.path.join(path, img), cv2.IMREAD_GRAYSCALE)
                new_array = cv2.resize(img_array, (IMG_SIZE, IMG_SIZE))
                training_data.append([new_array, class_num])
            except Exception as e:
                pass


create_training_data()

print(len(training_data))

import random

random.shuffle(training_data)
for sample in training_data[:10]:
    print(sample[1])

X = []
y = []
for features, label in training_data:
    X.append(features)
    y.append(label)

X = np.array(X).reshape(-1, IMG_SIZE, IMG_SIZE, 1)

import pickle

pickle_in = open("X.pickle","rb")
X = pickle.load(pickle_in)



pickle_in = open("y.pickle","rb")
y = pickle.load(pickle_in)

X = X/255.0

model = Sequential()

model.add(Conv2D(64, (3, 3), input_shape=X.shape[1:]))
model.add(Activation('relu'))
padding='same'
model.add(Conv2D(64, (3, 3), input_shape=X.shape[1:]))
model.add(Activation('relu'))
padding='same'
model.add(MaxPooling2D(pool_size=(1, 1)))

model.add(Conv2D(128, (3, 3), input_shape=X.shape[1:]))
model.add(Activation('relu'))
padding='same'
model.add(Conv2D(128, (3, 3), input_shape=X.shape[1:]))
model.add(Activation('relu'))
padding='same'
model.add(MaxPooling2D(pool_size=(1, 1)))

model.add(Conv2D(256, (3, 3), input_shape=X.shape[1:]))
model.add(Activation('relu'))
padding='same'
model.add(Conv2D(256, (3, 3), input_shape=X.shape[1:]))
model.add(Activation('relu'))
padding='same'
model.add(Conv2D(256, (3, 3), input_shape=X.shape[1:]))
model.add(Activation('relu'))
padding='same'
model.add(MaxPooling2D(pool_size=(1, 1)))

model.add(Conv2D(512, (3, 3), input_shape=X.shape[1:]))
model.add(Activation('relu'))
padding='same'
model.add(Conv2D(512, (3, 3), input_shape=X.shape[1:]))
model.add(Activation('relu'))
padding='same'
model.add(Conv2D(512, (3, 3), input_shape=X.shape[1:]))
model.add(Activation('relu'))
padding='same'
model.add(MaxPooling2D(pool_size=(1, 1)))

model.add(Conv2D(512, (3, 3), input_shape=X.shape[1:]))
model.add(Activation('relu'))
padding='same'
model.add(Conv2D(512, (3, 3), input_shape=X.shape[1:]))
model.add(Activation('relu'))
padding='same'
model.add(Conv2D(512, (3, 3), input_shape=X.shape[1:]))
model.add(Activation('relu'))
padding='same'
model.add(MaxPooling2D(pool_size=(1, 1)))

model.add(Flatten())

model.add(Dense(64))
model.add(Activation('relu'))

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

model.compile(loss='binary_crossentropy',
              optimizer='adam',
              metrics=['accuracy'])
model.fit(X, y, batch_size=15, epochs=1, validation_split=0.1)

from sklearn.metrics import confusion_matrix
pred = model.predict(X)
pred = np.round(pred)

conf = confusion_matrix(y, pred)

import seaborn as sns
sns.heatmap(conf, annot=True)

plt.show()

但是，运行这个总是给我一个57.69%的val准确性在任何时代，我做错了什么吗？还是我什么都做错了？

编辑，所以我现在使用了一个预先训练过的模型

import tensorflow as tf
from keras.datasets import cifar10
from keras.preprocessing.image import ImageDataGenerator
from keras.models import Sequential
from keras.layers import Dense, Dropout, Activation, Flatten
from keras.layers import Conv2D, MaxPooling2D
import numpy as np
import matplotlib.pyplot as plt
import os
import cv2

DATADIR = "C:/Users/Acer/imagerec/MRI"

CATEGORIES = ["yes", "no"]

for category in CATEGORIES:
    path = os.path.join(DATADIR,category)
    for img in os.listdir(path):
        img_array = cv2.imread(os.path.join(path,img) ,cv2.IMREAD_GRAYSCALE)
        plt.imshow(img_array, cmap='gray')


        break
    break
print(img_array)
print(img_array.shape)

IMG_SIZE = 224

new_array = cv2.resize(img_array, (IMG_SIZE, IMG_SIZE))
plt.imshow(new_array, cmap='gray')


training_data = []

def create_training_data():
    for category in CATEGORIES:
        path = os.path.join(DATADIR, category)
        class_num = CATEGORIES.index(category)
        for img in os.listdir(path):
            try:
                img_array = cv2.imread(os.path.join(path, img), cv2.IMREAD_GRAYSCALE)
                new_array = cv2.resize(img_array, (IMG_SIZE, IMG_SIZE))
                training_data.append([new_array, class_num])
            except Exception as e:
                pass


create_training_data()

print(len(training_data))

import random

random.shuffle(training_data)
for sample in training_data[:10]:
    print(sample[1])

X = []
y = []
for features, label in training_data:
    X.append(features)
    y.append(label)

X = np.array(X).reshape(-1, IMG_SIZE, IMG_SIZE, 1)

import pickle

pickle_in = open("X.pickle","rb")
X = pickle.load(pickle_in)



pickle_in = open("y.pickle","rb")
y = pickle.load(pickle_in)

X = X/255.0


def input_shape(args):
    pass


from keras.applications.vgg16 import VGG16
from keras.models import Model
from keras.layers import Dense

vgg = VGG16(include_top=False, weights='imagenet', input_shape=(), pooling='avg')
x = vgg.output
x = Dense(1, activation='sigmoid')(x)
model = Model(vgg.input, x)
model.summary()

from sklearn.metrics import confusion_matrix
pred = model.predict(X)
pred = np.round(pred)

conf = confusion_matrix(y, pred)

import seaborn as sns
sns.heatmap(conf, annot=True)

plt.show()

model.save('64x2-CNN.model')

得到了一个错误模型："model_1“

图层(类型)输出形状参数

input_1 (InputLayer) (无，3) 0

block1_conv1 (Conv2D) (无，64) 1792

block1_conv2 (Conv2D) (无，64) 36928

block1_pool (MaxPooling2D) (无，64) 0

block2_conv1 (Conv2D) (无，128) 73856

block2_conv2 (Conv2D) (无，128) 147584

block2_pool (MaxPooling2D) (无，128) 0

block3_conv1 (Conv2D) (无，256个) 295168

block3_conv2 (Conv2D) (无，256个) 590080

block3_conv3 (Conv2D) (无，256个) 590080

block3_pool (MaxPooling2D) (无，256个)0

block4_conv1 (Conv2D) (无，512) 1180160

block4_conv2 (Conv2D) (无，512) 2359808

block4_conv3 (Conv2D) (无，512) 2359808

block4_pool (MaxPooling2D) (无，512) 0

block5_conv1 (Conv2D) (无，512) 2359808

block5_conv2 (Conv2D) (无，512) 2359808

block5_conv3 (Conv2D) (无，512) 2359808

block5_pool (MaxPooling2D) (无，512) 0

global_average_pooling2d_1 (无，512) 0

dense_1 (稠密)(无，1) 513

助理人员总数: 14,715,201名可培训对讲员: 14,715,201名不可培训的副手:0

追溯(最近一次调用)：文件"C:/Users/Acer/PycharmProjects/condas/UwU.py"，第95行，在pred = model.predict(X)文件"C:\Users\Acer\Anaconda3\envs\condas\lib\site-packages\keras\engine\training.py"，行1441中，在预测x，_中，_= self._standardize_user_data(x)文件"C:\Users\Acer\Anaconda3\envs\condas\lib\site-packages\keras\engine\training.py"，行579，在"C:\Users\Acer\Anaconda3\envs\condas\lib\site-packages\keras\engine\training_utils.py"，_ _standardize_user_data _prefix=‘input’)文件的第145行中，在standardize_input_data str(Data_shape)中，ValueError:在检查输入时出错:期望input_1具有形状(无，无，3)，但有形状数组(50，50，1)

进程已完成，退出代码为%1

Answer 1

在keras序列模型中，只有第一个层需要知道它应该期望的input_shape，在您的例子中，它是Conv2D层。此外，添加多个乙状结肠激活的Dense层是没有意义的。

请参阅此

model = Sequential([
Conv2D(64, (3, 3), input_shape=input_shape, padding='same', activation='relu'),
Conv2D(64, (3, 3), activation='relu', padding='same'),
MaxPooling2D(pool_size=(2, 2), strides=(2, 2)),
Conv2D(128, (3, 3), activation='relu', padding='same'),
Conv2D(128, (3, 3), activation='relu', padding='same',),
MaxPooling2D(pool_size=(2, 2), strides=(2, 2)),
Conv2D(256, (3, 3), activation='relu', padding='same',),
Conv2D(256, (3, 3), activation='relu', padding='same',),
Conv2D(256, (3, 3), activation='relu', padding='same',),
MaxPooling2D(pool_size=(2, 2), strides=(2, 2)),
Conv2D(512, (3, 3), activation='relu', padding='same',),
Conv2D(512, (3, 3), activation='relu', padding='same',),
Conv2D(512, (3, 3), activation='relu', padding='same',),
MaxPooling2D(pool_size=(2, 2), strides=(2, 2)),
Conv2D(512, (3, 3), activation='relu', padding='same',),
Conv2D(512, (3, 3), activation='relu', padding='same',),
Conv2D(512, (3, 3), activation='relu', padding='same',),
MaxPooling2D(pool_size=(2, 2), strides=(2, 2)),
Flatten(),
Dense(4096, activation='relu'),
Dense(4096, activation='relu'),
Dense(1, activation='sigmoid')
])

或者，您可以从keras应用程序中使用经过预先训练的VGG模型。

from keras.applications.vgg16 import VGG16
from keras.models import Model
from keras.layers import Dense

vgg = VGG16(include_top=False, weights='imagenet', input_shape=(), pooling='avg')
x = vgg.output
x = Dense(1, activation='sigmoid')(x)
model = Model(vgg.input, x)
model.summary()