-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathModel.py
191 lines (151 loc) · 7.03 KB
/
Model.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
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
import numpy as np
import cv2
import os
import matplotlib.pyplot as plt
import tensorflow as tf
from sklearn.model_selection import train_test_split
from keras.models import load_model
from keras.losses import categorical_crossentropy
from sklearn.metrics import accuracy_score, precision_score, recall_score, f1_score
images = []
labels_str = []
def listeleme(kok_dizin):
for sinif in os.listdir(kok_dizin):
print(sinif)
sinif_dizini = os.path.join(kok_dizin, sinif)
if os.path.isdir(sinif_dizini):
for dosya in os.listdir(sinif_dizini):
if dosya.endswith(".png"):
yol = os.path.join(sinif_dizini, dosya)
img = cv2.imread(yol)
img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
img = img / 255
images.append(img)
labels_str.append(sinif)
kok_dizin = "Processed_Spectrograms"
listeleme(kok_dizin)
numaralandirma = {
"air_conditioner": 0,
"car_horn": 1,
"children_playing": 2,
"dog_bark": 3,
"drilling": 4,
"engine_idling": 5,
"gun_shot": 6,
"jackhammer": 7,
"siren": 8,
"street_music": 9
}
labels = [numaralandirma.get(oges, None) for oges in labels_str]
images = np.array(images)
labels = np.array(labels)
X_train, X_, y_train, y_ = train_test_split(images, labels, train_size=0.7, random_state=42)
X_test, X_val, y_test, y_val = train_test_split(X_, y_, test_size=0.5, random_state=42)
#MODEL
model_1 = tf.keras.models.Sequential()
model_1.add(tf.keras.layers.Conv2D(filters=32, kernel_size=3, activation='relu',
input_shape=[100, 100, 1]))
model_1.add(tf.keras.layers.MaxPool2D(pool_size=(2, 2)))
model_1.add(tf.keras.layers.Conv2D(filters=32, kernel_size=3, activation='relu'))
model_1.add(tf.keras.layers.MaxPool2D(pool_size=(2, 2)))
model_1.add(tf.keras.layers.Flatten())
model_1.add(tf.keras.layers.Dense(128, activation='relu'))
model_1.add(tf.keras.layers.Dense(64, activation='relu'))
model_1.add(tf.keras.layers.Dense(10, activation='softmax'))
model_1.compile(optimizer="adam", loss='sparse_categorical_crossentropy', metrics=['accuracy'])
model_2 = tf.keras.models.Sequential()
model_2.add(tf.keras.layers.Conv2D(filters=64, kernel_size=3, activation='relu',
input_shape=[100, 100, 1]))
model_2.add(tf.keras.layers.MaxPool2D(pool_size=(2, 2)))
model_2.add(tf.keras.layers.Conv2D(filters=32, kernel_size=3, activation='relu'))
model_2.add(tf.keras.layers.MaxPool2D(pool_size=(2, 2)))
model_2.add(tf.keras.layers.Flatten())
model_2.add(tf.keras.layers.Dense(128, activation='relu'))
model_2.add(tf.keras.layers.Dense(64, activation='relu'))
model_2.add(tf.keras.layers.Dense(10, activation='softmax'))
model_2.compile(optimizer="adam", loss='sparse_categorical_crossentropy', metrics=['accuracy'])
model_3 = tf.keras.models.Sequential()
model_3.add(tf.keras.layers.Conv2D(filters=32, kernel_size=3, activation='relu',
input_shape=[100, 100, 1]))
model_3.add(tf.keras.layers.MaxPool2D(pool_size=(2, 2)))
model_3.add(tf.keras.layers.Conv2D(filters=32, kernel_size=3, activation='relu'))
model_3.add(tf.keras.layers.MaxPool2D(pool_size=(2, 2)))
model_3.add(tf.keras.layers.Flatten())
model_3.add(tf.keras.layers.Dense(128, activation='relu'))
model_3.add(tf.keras.layers.Dense(10, activation='softmax'))
model_3.compile(optimizer="adam", loss='sparse_categorical_crossentropy', metrics=['accuracy'])
model_4 = tf.keras.models.Sequential()
model_4.add(tf.keras.layers.Conv2D(filters=64, kernel_size=3, activation='relu',
input_shape=[100, 100, 1]))
model_4.add(tf.keras.layers.MaxPool2D(pool_size=(2, 2)))
model_4.add(tf.keras.layers.Conv2D(filters=32, kernel_size=3, activation='relu'))
model_4.add(tf.keras.layers.MaxPool2D(pool_size=(2, 2)))
model_4.add(tf.keras.layers.Flatten())
model_4.add(tf.keras.layers.Dense(128, activation='relu'))
model_4.add(tf.keras.layers.Dense(10, activation='softmax'))
model_4.compile(optimizer="adam", loss='sparse_categorical_crossentropy', metrics=['accuracy'])
accuracy_score_train = []
loss_train = []
accuracy_score_val = []
loss_val = []
models = [model_1, model_2, model_3, model_4]
for model in models:
tmp = model.fit(X_train, y_train, epochs=10, validation_data=(X_val, y_val))
accuracy_score_train.append(tmp.history['accuracy'])
loss_train.append(tmp.history['loss'])
accuracy_score_val.append(tmp.history['val_accuracy'])
loss_val.append(tmp.history['val_loss'])
accuracy_score_train = np.array(accuracy_score_train)
loss_train = np.array(loss_train)
accuracy_score_val = np.array(accuracy_score_val)
loss_val = np.array(loss_val)
plt.rcParams.update({'font.size': 10})
fig, axes = plt.subplots(nrows=4, ncols=2, figsize=(16, 12))
for i, ax in enumerate(axes.flat):
if i % 2 == 0:
ax.plot(accuracy_score_train[i // 2])
ax.plot(accuracy_score_val[i // 2])
ax.set_title(f'Model {i // 2 + 1} Accuracy')
ax.set_xlabel('Epochs')
ax.set_ylabel('Accuracy')
ax.set_xticks(range(10))
ax.set_xticklabels(np.arange(1, 11))
ax.legend(['Train', 'Validation'])
else:
ax.plot(loss_train[i // 2])
ax.plot(loss_val[i // 2])
ax.set_title(f'Model {i // 2 + 1} Loss')
ax.set_xlabel('Epochs')
ax.set_ylabel('Loss')
ax.set_xticks(range(10))
ax.set_xticklabels(np.arange(1, 11))
ax.legend(['Train', 'Validation'])
img = plt.tight_layout()
plt.savefig('metric_graph.png')
plt.show()
model = tf.keras.models.Sequential()
model.add(tf.keras.layers.Conv2D(filters=32, kernel_size=3, activation='relu',
input_shape=[100, 100, 1]))
model.add(tf.keras.layers.MaxPool2D(pool_size=(2, 2)))
model.add(tf.keras.layers.Conv2D(filters=32, kernel_size=3, activation='relu'))
model.add(tf.keras.layers.MaxPool2D(pool_size=(2, 2)))
model.add(tf.keras.layers.Flatten())
model.add(tf.keras.layers.Dense(128, activation='relu'))
model.add(tf.keras.layers.Dense(64, activation='relu'))
model.add(tf.keras.layers.Dense(10, activation='softmax'))
model.compile(optimizer="adam", loss='sparse_categorical_crossentropy', metrics=['accuracy'])
model.fit(X_train, y_train, epochs=7, validation_data=(X_val, y_val))
model.save('cnn_model.h5') # Sürekli Sürekli Eğitmemek için kaydettik.
"""
loaded_model = load_model('cnn_model.h5')
"""
predictions = model.predict(X_test)
predicted_classes = np.argmax(predictions, axis=1)
print(f"Accuracy: {round(accuracy_score(y_test, predicted_classes), 2)}")
print(f"Recall: {round(recall_score(y_test,predicted_classes, average='weighted'),3)}")
print(f"Precision: {round(precision_score(y_test,predicted_classes, average='weighted'), 2)}")
print(f"F1: {round(f1_score(y_test,predicted_classes, average='weighted'), 2)}")
#Accuracy: 0.87
#Recall: 0.866
#Precision: 0.87
#F1: 0.87