-
Notifications
You must be signed in to change notification settings - Fork 0
/
sign_language.py
236 lines (168 loc) · 5.11 KB
/
sign_language.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
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
#Sign Language
# To add a new cell, type '# %%'
# To add a new markdown cell, type '# %% [markdown]'
# %%
# standart data tools
import numpy as np
import pandas as pd
# common visualizing tools
import matplotlib.pyplot as plt
import seaborn as sns
# CNN layers and the Deep Learning model
from tensorflow import keras
from keras.models import Sequential
from keras.layers import Conv2D, MaxPooling2D, Dense , Flatten, Dropout
from keras.optimizers import Adam
# splitting tool for the validation set
from sklearn.model_selection import train_test_split
# %%
X = np.load('X.npy')
Y = np.load('Y.npy')
# %%
print(" Max value of X: ",X.max())
print(" Min value of X: ",X.min())
print(" Shape of X: ",X.shape)
print("\n Max value of Y: ",Y.max())
print(" Min value of Y: ",Y.min())
print(" Shape of Y: ",Y.shape)
# %%
plt.figure(figsize=(24,8))
plt.subplot(2,5,1)
plt.title(Y[0].argmax())
plt.imshow(X[0])
plt.axis("off")
plt.subplot(2,5,2)
plt.title(Y[1000].argmax())
plt.imshow(X[1000])
plt.axis("off")
plt.subplot(2,5,3)
plt.title(Y[222].argmax())
plt.imshow(X[222])
plt.axis("off")
plt.subplot(2,5,4)
plt.title(Y[1500].argmax())
plt.imshow(X[1500])
plt.axis("off")
plt.subplot(2,5,5)
plt.title(Y[2000].argmax())
plt.imshow(X[2000])
plt.axis("off")
plt.subplot(2,5,6)
plt.title(Y[1200].argmax())
plt.imshow(X[1200])
plt.axis("off")
plt.subplot(2,5,7)
plt.title(Y[1700].argmax())
plt.imshow(X[1700])
plt.axis("off")
plt.subplot(2,5,8)
plt.title(Y[500].argmax())
plt.imshow(X[500])
plt.axis("off")
plt.subplot(2,5,9)
plt.title(Y[700].argmax())
plt.imshow(X[700])
plt.axis("off")
plt.subplot(2,5,10)
plt.title(Y[1400].argmax())
plt.imshow(X[1400])
plt.axis("off")
plt.suptitle("Example of each sign", fontsize=20)
plt.show()
# %%
x_9 = X[:204]; x_0 = X[204:409]; x_7 = X[409:615]; x_6 = X[615:822]; x_1 = X[822:1028];
x_8 = X[1028:1236]; x_4 = X[1236:1443]; x_3 = X[1443:1649]; x_2 = X[1649:1855]; x_5 = X[1855:];
# %%
mean0 = x_0.mean(axis=0); mean1 = x_1.mean(axis=0); mean2 = x_2.mean(axis=0);
mean3 = x_3.mean(axis=0); mean4 = x_4.mean(axis=0); mean5 = x_5.mean(axis=0);
mean6 = x_6.mean(axis=0); mean7 = x_7.mean(axis=0); mean8 = x_8.mean(axis=0); mean9 = x_9.mean(axis=0);
# %%
plt.figure(figsize=(24,8))
plt.subplot(2,5,1)
plt.title("Mean of digit 0")
plt.imshow(mean0)
plt.axis("off")
plt.subplot(2,5,2)
plt.title("Mean of digit 1")
plt.imshow(mean1)
plt.axis("off")
plt.subplot(2,5,3)
plt.title("Mean of digit 2")
plt.imshow(mean2)
plt.axis("off")
plt.subplot(2,5,4)
plt.title("Mean of digit 3")
plt.imshow(mean3)
plt.axis("off")
plt.subplot(2,5,5)
plt.title("Mean of digit 4")
plt.imshow(mean4)
plt.axis("off")
plt.subplot(2,5,6)
plt.title("Mean of digit 5")
plt.imshow(mean5)
plt.axis("off")
plt.subplot(2,5,7)
plt.title("Mean of digit 6")
plt.imshow(mean6)
plt.axis("off")
plt.subplot(2,5,8)
plt.title("Mean of digit 7")
plt.imshow(mean7)
plt.axis("off")
plt.subplot(2,5,9)
plt.title("Mean of digit 8")
plt.imshow(mean8)
plt.axis("off")
plt.subplot(2,5,10)
plt.title("Mean of digit 9")
plt.imshow(mean9)
plt.axis("off")
plt.suptitle("Mean of each sign", fontsize=20)
plt.show()
# %%
x_train, x_test, y_train, y_test = train_test_split(X,Y,test_size=0.2,random_state=42)
x_train = x_train.reshape(-1,64,64,1)
x_test = x_test.reshape(-1,64,64,1)
# %%
CNN_model = Sequential()
CNN_model.add(Conv2D(filters=16,kernel_size=(5,5),activation="relu",padding="same",input_shape=(64,64,1)))
CNN_model.add(Conv2D(filters=32,kernel_size=(5,5),activation="relu",padding="same"))
CNN_model.add(MaxPooling2D(pool_size=(2,2),padding="same"))
CNN_model.add(Conv2D(filters=32,kernel_size=(4,4),activation="relu",padding="same"))
CNN_model.add(Conv2D(filters=32,kernel_size=(4,4),activation="relu",padding="same"))
CNN_model.add(MaxPooling2D(pool_size=(2,2),padding="same"))
CNN_model.add(Dropout(0.2))
CNN_model.add(Conv2D(filters=32,kernel_size=(3,3),activation="relu",padding="same"))
CNN_model.add(Conv2D(filters=64,kernel_size=(3,3),activation="relu",padding="same"))
CNN_model.add(MaxPooling2D(pool_size=(2,2),padding="same"))
CNN_model.add(Dropout(0.2))
CNN_model.add(Conv2D(filters=32,kernel_size=(2,2),activation="relu",padding="same"))
CNN_model.add(MaxPooling2D(pool_size=(2,2),padding="same"))
CNN_model.add(Dropout(0.2))
CNN_model.add(Flatten())
CNN_model.add(Dense(128,activation="relu"))
CNN_model.add(Dense(64,activation="relu"))
CNN_model.add(Dense(40,activation="relu"))
CNN_model.add(Dense(10,activation="softmax"))
# %%
CNN_model.summary()
# %%
CNN_model.compile(optimizer=Adam(lr=0.0002),loss=keras.losses.categorical_crossentropy,metrics=["accuracy"])
# %%
results = CNN_model.fit(x_train,y_train,epochs=70,validation_data=(x_test,y_test))
# %%
plt.figure(figsize=(24,8))
plt.subplot(1,2,1)
plt.plot(results.history["val_accuracy"],label="validation_accuracy",c="red",linewidth=4)
plt.plot(results.history["accuracy"],label="training_accuracy",c="green",linewidth=4)
plt.legend()
plt.grid(True)
plt.subplot(1,2,2)
plt.plot(results.history["val_loss"],label="validation_loss",c="red",linewidth=4)
plt.plot(results.history["loss"],label="training_loss",c="green",linewidth=4)
plt.legend()
plt.grid(True)
plt.suptitle("ACC / LOSS",fontsize=18)
plt.show()
# %%