-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathsift_bow_tf_svm_final.py
147 lines (106 loc) · 4.3 KB
/
sift_bow_tf_svm_final.py
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
import cv2
import numpy as np
from mnist import MNIST
from sklearn.cluster import KMeans
import time
from sklearn.neighbors import NearestNeighbors
start_time = time.time()
k1 = 5
k2 = 6
mndata = MNIST('./MNIST/')
# Train
(X_train, labels_train) = mndata.load_training()
X_train = np.array(X_train, dtype='uint8')
labels_train = np.array(labels_train)
X_train_k1 = X_train[labels_train == k1, :]
X_train_k2 = X_train[labels_train == k2, :]
X_train = np.concatenate((X_train_k1, X_train_k2), axis=0)
labels_train = np.concatenate(([k1] * X_train_k1.shape[0], [k2] * X_train_k2.shape[0]), axis=0)
descriptors = []
sift = cv2.xfeatures2d.SIFT_create()
des_keypoint_train_final = []
labels_train_final = []
for i in range(labels_train.shape[0]):
pixels = X_train[i, :].reshape((28, 28))
kp_mnist, des_mnist = sift.detectAndCompute(pixels, None)
if des_mnist is not None:
des_keypoint_train_final.append(des_mnist)
labels_train_final.append(labels_train[i])
des_mnist = des_mnist.tolist()
for des in des_mnist:
descriptors.append(des)
descriptors = np.array(descriptors)
# descriptors = descriptors / np.max(descriptors)
# Tao tu dien
K = 100
kmeans = KMeans(n_clusters=K).fit(descriptors)
bow = kmeans.cluster_centers_.tolist()
# Dua anh ve vector theo bow
X_train_final = []
neigh = NearestNeighbors(5).fit(bow)
for des_keypoint in des_keypoint_train_final:
x = np.array([0] * K)
for i in des_keypoint:
dist, nearest_id = neigh.kneighbors([i], 1)
x[nearest_id] += 1
X_train_final.append(x)
X_train_final = np.array(X_train_final)
# Test_data
(X_test, labels_test) = mndata.load_testing()
X_test = np.array(X_test, dtype='uint8')
labels_test = np.array(labels_test)
X_test_k1 = X_test[labels_test == k1, :]
X_test_k2 = X_test[labels_test == k2, :]
X_test = np.concatenate((X_test_k1, X_test_k2), axis=0)
labels_test = np.concatenate(([k1] * X_test_k1.shape[0], [k2] * X_test_k2.shape[0]), axis=0)
labels_test = np.array([1 if label == k1 else -1 for label in labels_test])
des_keypoint_test_final = []
labels_test_final = []
for i in range(labels_test.shape[0]):
pixels = X_test[i, :].reshape((28, 28))
kp_mnist, des_mnist = sift.detectAndCompute(pixels, None)
if des_mnist is not None:
des_keypoint_test_final.append(des_mnist)
labels_test_final.append(labels_test[i])
labels_test_final = np.array(labels_test_final)
X_test_final = []
for des_keypoint in des_keypoint_test_final:
x = np.array([0] * K)
for i in des_keypoint:
dist, nearest_id = neigh.kneighbors([i], 1)
x[nearest_id] += 1
X_test_final.append(x)
X_test_final = np.array(X_test_final)
# Train
import tensorflow as tf
labels_train_final = np.array([1 if y == k1 else -1 for y in labels_train_final])
x_data = tf.placeholder(shape=[None, K], dtype=tf.float32)
y_target = tf.placeholder(shape=[None, 1], dtype=tf.float32)
print(x_data, y_target)
W = tf.Variable(tf.random_normal(shape=[K, 1]))
b = tf.Variable(tf.random_normal(shape=[1, 1]))
model_output = tf.add(tf.matmul(x_data, W), b)
l2_norm = tf.norm(W)
hinge = tf.reduce_sum(tf.maximum(0., tf.subtract(1., tf.multiply(model_output, y_target))))
loss = hinge + l2_norm
optimizer = tf.train.GradientDescentOptimizer(learning_rate=0.001).minimize(loss)
labels_train_final = labels_train_final.reshape(labels_train_final.shape[0], 1)
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
for i in range(500):
_, loss_ = sess.run([optimizer, loss], feed_dict={x_data: X_train_final, y_target: labels_train_final})
W_out, b_out = sess.run([W, b])
out_test = sess.run(model_output, feed_dict={x_data: X_test_final})
labels_predict = np.array([1 if out > 0 else -1 for out in out_test])
count = 0.0
for i in range(labels_predict.shape[0]):
if labels_predict[i] != labels_test_final[i]:
count += 1.0
print(count, '===', labels_predict.shape[0])
print(100.0 - 100 * count / labels_predict.shape[0])
print('W = ', W_out.T, '\n=======\n', 'b = ', b_out)
end_train = time.time()
print(end_train - start_time)
f = open('./W_b.txt', 'w+')
f.write('%s \n' % W_out)
f.write('%s \n' % b_out)