image-multi-task-learning-/single_task_cnn.py at master · Gaelic98/image-multi-task-learning- · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150

# coding: utf-8

from __future__ import print_function
import keras
from keras.datasets import cifar10, mnist
from keras.models import Sequential, Model
from keras.layers import Dense, Dropout, Activation, Flatten, Input
from keras.layers import Conv2D, MaxPooling2D,GlobalAveragePooling2D,GlobalAveragePooling1D
from keras.layers.advanced_activations import LeakyReLU
from keras.layers.normalization import BatchNormalization
from keras.callbacks import ModelCheckpoint, TensorBoard
from keras.initializers import RandomUniform
from keras.utils import plot_model
import os
import pickle
import numpy as np
import tensorflow as tf

from img_process import *
from lsuv_init import LSUVinit
import mnist_reader

flags = tf.app.flags
flags.DEFINE_integer("epochs", 25, "Epoch to train")
flags.DEFINE_float("learning_rate", 0.001, "Learning rate of for adam[0.001] ")
flags.DEFINE_boolean("load_model", False, "loading model weights")
flags.DEFINE_boolean("multi_task", False, "using pretraning classifier weights")
flags.DEFINE_integer("datasets", 0, "0:using MNIST datasets,1:using fashionMNIST datasets,2:using notMNIST datasets")
FLAGS = flags.FLAGS
num_classes = 10
batch_size = 100

if(FLAGS.datasets == 0):
    data_class_name=['0','1','2','3','4','5','6','7','8','9']
    (x_train, y_train), (x_test, y_test) = mnist.load_data()
if(FLAGS.datasets == 1):
    data_class_name=['T-shirt','Trouser','Pullover','Dress','Coat','Sandal','Shirt','Sneaker','Bag','Ankle boot']
    x_train, y_train = mnist_reader.load_mnist('data/fashion', kind='train')
    x_test, y_test = mnist_reader.load_mnist('data/fashion', kind='t10k')
if(FLAGS.datasets == 2):
    data_class_name=['A','B','C','D','E','F','G','H','I','J']
    x_train, y_train = mnist_reader.load_mnist('data/notMNIST', kind='train')
    x_test, y_test = mnist_reader.load_mnist('data/notMNIST', kind='t10k')

print('x_train shape:', x_train.shape)
x_train=re_scale(x_train)
x_test=re_scale(x_test)#重塑至[0,1]

x_train = np.reshape(x_train,[-1,28,28,1])
x_test = np.reshape(x_test,[-1,28,28,1])
x_train = x_train.astype('float32')
x_test = x_test.astype('float32')
#x_train=i2b(x_train)
#x_test=i2b(x_test)#image to binary

# Convert class vectors to binary class matrices.
y_train = keras.utils.to_categorical(y_train, num_classes)
y_test = keras.utils.to_categorical(y_test, num_classes)

# The data, shuffled and split between train and test sets:
print('x_train shape:', x_train.shape)
print(y_train.shape)
print(x_train.shape[0], 'train samples')
print(x_test.shape[0], 'test samples')

model = Sequential()
model.add(Conv2D(96, (3, 3), padding='same',input_shape=x_train.shape[1:]))
model.add(BatchNormalization())
model.add(Activation('relu'))
model.add(Conv2D(96, (3, 3), padding='same'))
model.add(BatchNormalization())
model.add(Activation('relu'))
model.add(Conv2D(96, (3, 3), padding='same',strides=(2,2)))
model.add(BatchNormalization())
model.add(Activation('relu'))
model.add(Dropout(0.5))

model.add(Conv2D(192, (3, 3), padding='same'))
model.add(BatchNormalization())
model.add(Activation('relu'))
model.add(Conv2D(192, (3, 3), padding='same'))
model.add(BatchNormalization())
model.add(Activation('relu'))
model.add(Conv2D(192, (3, 3), padding='same',strides=(2,2)))
model.add(BatchNormalization())
model.add(Activation('relu'))
model.add(Dropout(0.5))

model.add(Conv2D(192, (3, 3), padding='valid'))
model.add(BatchNormalization())
model.add(Activation('relu'))

model.add(Conv2D(192, (1, 1), padding='same'))
model.add(BatchNormalization())
model.add(Activation('relu'))

model.add(Conv2D(10, (1, 1), padding='same'))
model.add(BatchNormalization())
model.add(Activation('relu'))
model.add(GlobalAveragePooling2D(data_format='channels_last'))
model.add(BatchNormalization())
model.add(Activation('softmax'))

model.summary()
#plot_model(model, to_file='model.png')
if(FLAGS.multi_task==True):
    for i in range(27):
        w = np.load('./layer_weight/weights{}.00.npy'.format(i))
        if(i==26):
            w[0] = w[0][:,:,:,0:10]
            w[1] = w[1][0:10]
        model.layers[i].set_weights(w)


# train the model
decay_rate = FLAGS.learning_rate / FLAGS.epochs
opt = keras.optimizers.Adam(lr=FLAGS.learning_rate,decay=decay_rate)
model.compile(loss='categorical_crossentropy',optimizer=opt,metrics=['accuracy'])
if (FLAGS.load_model):
    model.load_weights('./checkpoint/learning_decay_weights.h5')
model.fit(x_train, y_train,batch_size,epochs=FLAGS.epochs,validation_data=(x_test, y_test),verbose=2)
if (FLAGS.load_model == False):
    model.save_weights('./checkpoint/learning_decay_weights.h5')
model.save_weights('./checkpoint/cnn_weights.h5')

result = model.predict(x_test)

matrix = np.zeros((10, 10))
right=0

for i, (Target, Label) in enumerate( zip(y_test, result) ) :  ### i-th label
    m = np.max(Label)
    for j, value in enumerate(Label) :  ### find max value and position
        if value == m :
            for k, num in enumerate(Target) :  ### find test calss
                if num == 1 :
                    matrix[k][j] += 1
                    break  # end of for k
            break  # end of for j

print('classification result')
for i in range(10):
    print('{}:{:.2f}% '.format(data_class_name[i],(matrix[i][i]/np.sum(matrix[i]))*100),end="")
    if i==4:print()
    right=right+matrix[i][i]

print("total:",len(result))
print("correct number：",right)
print("correct rate：{:.2f}%".format(right/len(result)*100))