-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathModelANNTensorflow.py
233 lines (173 loc) · 8.13 KB
/
ModelANNTensorflow.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
import os
'''trainModelsFunctions'''
def trainModelWithOnlyACC(threshold, numMin, numMax,preprocessingData):
import tensorflow.compat.v2.feature_column as fc
import tensorflow as tf
from tensorflow import keras
from tensorflow.keras import optimizers
from os.path import join
from pathlib import Path
"""**ANN Modeling**"""
labelTrain, labelTest, MyNewDataSetTrain, MyNewDataSetTest = preprocessingData
##cria e compila o modelo da rede neural
model = keras.Sequential([
keras.layers.Dense(9), # input layer (1)
keras.layers.Dense(120, activation='relu'), # hidden layer (2)
keras.layers.Dense(120, activation='relu'), # hidden layer (2)
keras.layers.Dense(30, activation='softmax') # output layer (3)
])
##Treina o modelo
cont = 0
valueAccuracy = 0
resultModel = None
while cont < numMin or (valueAccuracy < threshold and cont < numMax):
model.compile(optimizer='adam',
loss='sparse_categorical_crossentropy',
metrics=['accuracy'])
model.fit(MyNewDataSetTrain, labelTrain, epochs=300) # we pass the data, labels and epochs and watch the magic!
##Avalia o modelo
test_loss, test_acc = model.evaluate(MyNewDataSetTest, labelTest, verbose=1)
if valueAccuracy < test_acc:
valueAccuracy = test_acc
resultModel = model
cont+=1
##Avalia o modelo
#test_loss, test_acc = model.evaluate(MyNewDataSetTest, labelTest, verbose=1)
print('Test accuracy:', test_acc)
"""**Export Model**"""
tms_model = resultModel.save('saved_model/my_model_')
"""**Load Model**"""
#path = os.path.join('saved_model/my_model_')
#loaded = tf.keras.models.load_model(path)
#print('Test accuracy:', test_acc)
"""**Generate Model Trained File**"""
##pip install -q tflite_support
##**Transfort to Tensorflow lite model**
_TFLITE_MODEL_PATH = "saved_model/"+"ANNModel2.tflite"
converter = tf.lite.TFLiteConverter.from_saved_model('saved_model/my_model_')
converter.optimizations = [tf.lite.Optimize.DEFAULT]
tflite_model = converter.convert()
with open(_TFLITE_MODEL_PATH, 'wb') as f:
f.write(tflite_model)
PATH_DIR = Path.cwd()
dataset_dir = PATH_DIR.joinpath('saved_model')
tflite_model_file = dataset_dir.joinpath("ANNModel2.tflite"+'32')
tflite_model_file.write_bytes(tflite_model)
converter = tf.lite.TFLiteConverter.from_saved_model('saved_model/my_model_')
converter.optimizations = [tf.lite.Optimize.DEFAULT]
converter.target_spec.supported_types = [tf.float16]
tflite_model = converter.convert()
PATH_DIR = Path.cwd()
dataset_dir = PATH_DIR.joinpath('saved_model')
tflite_model_file = dataset_dir.joinpath("ANNModel2.tflite"+'16')
tflite_model_file.write_bytes(tflite_model)
return [resultModel,valueAccuracy]
def trainModelWithACCGYR(threshold, numMin, numMax,preprocessingData):
import tensorflow.compat.v2.feature_column as fc
import tensorflow as tf
from tensorflow import keras
from tensorflow.keras import optimizers
from os.path import join
from pathlib import Path
"""**ANN Modeling**"""
labelTrain, labelTest, MyNewDataSetTrain, MyNewDataSetTest = preprocessingData
##cria e compila o modelo da rede neural
model = keras.Sequential([
keras.layers.Dense(18), # input layer (1)
keras.layers.Dense(120, activation='relu'), # hidden layer (2)
keras.layers.Dense(120, activation='relu'), # hidden layer (2)
keras.layers.Dense(17, activation='softmax') # output layer (3)
])
##Treina o modelo
cont = 0
valueAccuracy = 0
resultModel = None
while cont < numMin or (valueAccuracy < threshold and cont < numMax):
model.compile(optimizer='adam',
loss='sparse_categorical_crossentropy',
metrics=['accuracy'])
model.fit(MyNewDataSetTrain, labelTrain, epochs=300) # we pass the data, labels and epochs and watch the magic!
##Avalia o modelo
test_loss, test_acc = model.evaluate(MyNewDataSetTest, labelTest, verbose=1)
if valueAccuracy < test_acc:
valueAccuracy = test_acc
resultModel = model
cont+=1
print('Test accuracy:', valueAccuracy)
"""**Export Model**"""
tms_model = resultModel.save('saved_model/my_model')
"""**Generate Model Trained File**"""
##pip install -q tflite_support
##**Transfort to Tensorflow lite model**
_TFLITE_MODEL_PATH = "saved_model/"+"ANNModel.tflite"
converter = tf.lite.TFLiteConverter.from_saved_model('saved_model/my_model')
converter.optimizations = [tf.lite.Optimize.DEFAULT]
tflite_model = converter.convert()
with open(_TFLITE_MODEL_PATH, 'wb') as f:
f.write(tflite_model)
PATH_DIR = Path.cwd()
dataset_dir = PATH_DIR.joinpath('saved_model')
tflite_model_file = dataset_dir.joinpath("ANNModel.tflite"+'32')
tflite_model_file.write_bytes(tflite_model)
converter = tf.lite.TFLiteConverter.from_saved_model('saved_model/my_model')
converter.optimizations = [tf.lite.Optimize.DEFAULT]
converter.target_spec.supported_types = [tf.float16]
tflite_model = converter.convert()
PATH_DIR = Path.cwd()
dataset_dir = PATH_DIR.joinpath('saved_model')
tflite_model_file = dataset_dir.joinpath("ANNModel.tflite"+'16')
tflite_model_file.write_bytes(tflite_model)
return [resultModel,valueAccuracy]
'''Predict'''
def predict(model, classes, correct_label):
import numpy as np
class_names = ['Andar', 'BATER_NA_MESA', 'BATER_PAREDE', 'CORRENDO', 'DEITAR', 'ESBARRAR_PAREDE', 'ESCREVER', 'PALMAS_EMP', 'PALMAS_SEN', 'PULO','QUEDA_APOIO_FRENTE', 'QUEDA_LATERAL', 'QUEDA_SAPOIO_FRENTE', 'SENTAR', 'SENTAR_APOIO', 'SENTAR_SAPOIO', 'TATEAR' ]
prediction = model.predict(np.array([classes]))
predicted_class = class_names[np.argmax(prediction)]
print(class_names[correct_label])
print(predicted_class)
def get_number():
while True:
num = input("Pick a number: ")
if num.isdigit():
num = int(num)
if 0 <= num <= 1000:
return int(num)
else:
print("Try again...")
if __name__ == '__main__':
from Dataset1 import Dataset1
from Dataset2 import Dataset2
from DatasetACC import DatasetACC
from DatasetACC_GYR import DatasetACC_GYR
print("====Preprocessing====")
dt1 = Dataset1("Datasets/Dataset1")
dt2 = Dataset2("Datasets/D2_ADL_Dataset/HMP_Dataset/All_data")
dtACC = DatasetACC([dt1,dt2])
dtACCGYR = DatasetACC_GYR([dt1])
ppDataACC = dtACC.executePreprocessing()
ppDataACCGYR = dtACCGYR.executePreprocessing()
print("====Training====")
modelGenerated2 = trainModelWithOnlyACC(0.7, 1, 10, ppDataACC)
modelGenerated = trainModelWithACCGYR(0.7, 1, 10, ppDataACCGYR)
print("=========Initialize Graph Generate Data===========")
algorithName = ["Artificial Neural Network"]
modelName = ["ANNModel.tflite"]
sensorList = [["acc", "accelerometer"], ["gyr","gyroscope"]]
featureList = ["acc_mean", "acc_max", "acc_min", "acc_std", "acc_kurtosis", "acc_skewness", "acc_entropy", "acc_mad", "acc_iqr","gyr_mean", "gyr_max", "gyr_min", "gyr_std", "gyr_kurtosis", "gyr_skewness", "gyr_entropy", "gyr_mad", "gyr_iqr"]
modelName2 = ["ANNModel2.tflite"]
sensorList2 = [["acc", "accelerometer"]]
featureList2 = ["acc_mean", "acc_max", "acc_min", "acc_std", "acc_kurtosis", "acc_skewness", "acc_entropy", "acc_mad", "acc_iqr"]
finalStateList = ['Andar', 'BATER_NA_MESA', 'BATER_PAREDE', 'CORRENDO', 'DEITAR','ESBARRAR_PAREDE', 'ESCREVER', 'PALMAS_EMP', 'PALMAS_SEN', 'PULO','QUEDA_APOIO_FRENTE', 'QUEDA_LATERAL', 'QUEDA_SAPOIO_FRENTE', 'SENTAR', 'SENTAR_APOIO', 'SENTAR_SAPOIO', 'TATEAR']
print("=========Predict===========")
labelTrain = ppDataACC[0]
labelTest = ppDataACC[1]
MyNewDataSetTrain = ppDataACC[2]
MyNewDataSetTest = ppDataACC[3]
while True:
num = get_number()
if num > 37:
break
classes = MyNewDataSetTest[num]
label = labelTest[num]
predict(modelGenerated2[0], classes, label)