ms762wu/EMG_Biometrics_2021
2
1
Fork 0
Code Issues Pull Requests Releases Wiki Activity

feat: add 1D convolution network in NNA

Browse Source
This commit is contained in:
Skudalen 2021-07-14 12:14:20 +02:00
parent 7ad034fa95
commit 968759d205
1 changed files with 23 additions and 25 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

48
Neural_Network_Analysis.py
View File

@ -241,13 +241,19 @@ def session_cross_validation(model_name:str, X, y, session_lengths, nr_sessions,
session_training_results = [] session_training_results = []
for i in range(nr_sessions): for i in range(nr_sessions):
X_train_session, X_test_session, y_train_session, y_test_session = prepare_datasets_sessions(X, y, session_lengths, i)
# Model: # Model:
if model_name == 'LSTM': if model_name == 'LSTM':
model = LSTM(input_shape=(1, 208)) model = LSTM(input_shape=(1, 208))
elif model_name == 'GRU': elif model_name == 'GRU':
model = GRU(input_shape=(1, 208)) model = GRU(input_shape=(1, 208))
elif model_name == 'CNN': elif model_name == 'CNN':
model = CNN(input_shape=(52, 52, 104)) print(X_train_session.shape)
print(X_test_session.shape)
X_train_session = np.reshape(X_train_session, (X_train_session.shape[0], 208, 1))
X_test_session = np.reshape(X_test_session, (X_test_session.shape[0], 208, 1))
model = CNN(input_shape=(208, 1))
elif model_name == 'FNN': elif model_name == 'FNN':
model = FFN(input_shape=(1, 208)) model = FFN(input_shape=(1, 208))
else: else:
@ -255,13 +261,7 @@ def session_cross_validation(model_name:str, X, y, session_lengths, nr_sessions,
model.summary() model.summary()
X_train_session, X_test_session, y_train_session, y_test_session = prepare_datasets_sessions(X, y, session_lengths, i)
if model_name == 'CNN':
X_train_session = X_train_session[..., np.newaxis]
X_test_session = X_test_session[..., np.newaxis]
X_train_session = np.reshape(X_train_session, (X_train_session.shape[0], 52, 52, 104))
X_test_session = np.reshape(X_test_session, (X_test_session.shape[0], 52, 52, 104))
train(model, X_train_session, y_train_session, verbose=1, batch_size=batch_size, epochs=epochs) train(model, X_train_session, y_train_session, verbose=1, batch_size=batch_size, epochs=epochs)
test_loss, test_acc = model.evaluate(X_test_session, y_test_session, verbose=2) test_loss, test_acc = model.evaluate(X_test_session, y_test_session, verbose=2)
session_training_results.append(test_acc) session_training_results.append(test_acc)
@ -326,21 +326,14 @@ def FFN(input_shape, nr_classes=5):
def CNN(input_shape, nr_classes=5): def CNN(input_shape, nr_classes=5):
model = keras.Sequential(name='CNN_model') model = keras.Sequential(name='CNN_model')
model.add(keras.layers.Conv2D(32, (3, 3), activation='relu', input_shape=input_shape)) #model.add(keras.layers.Input(name='the_input', shape=input_shape, dtype='float32'))
model.add(keras.layers.MaxPooling2D((3, 3), strides=(2, 2), padding='same')) model.add(keras.layers.Conv1D(32, kernel_size= 5, activation='relu', input_shape=input_shape)) # , input_shape=input_shape
model.add(keras.layers.BatchNormalization()) model.add(keras.layers.MaxPooling1D(pool_size=5))
#model.add(keras.layers.BatchNormalization())
model.add(keras.layers.Conv2D(32, (3, 3), activation='relu'))
model.add(keras.layers.MaxPooling2D((3, 3), strides=(2, 2), padding='same'))
model.add(keras.layers.BatchNormalization())
model.add(keras.layers.Conv2D(32, (2, 2), activation='relu'))
model.add(keras.layers.MaxPooling2D((2, 2), strides=(2, 2), padding='same'))
model.add(keras.layers.BatchNormalization())
# flatten output and feed it into dense layer
model.add(keras.layers.Flatten()) model.add(keras.layers.Flatten())
model.add(keras.layers.Dense(64, activation='relu')) #model.add(keras.layers.GlobalAveragePooling1D())
model.add(keras.layers.Dense(64, activation='relu')) # , input_shape=(...,1)
model.add(keras.layers.Dropout(0.3)) model.add(keras.layers.Dropout(0.3))
# Ouput layer # Ouput layer
model.add(keras.layers.Dense(nr_classes, activation='softmax', name='Softmax')) model.add(keras.layers.Dense(nr_classes, activation='softmax', name='Softmax'))
@ -371,14 +364,16 @@ if __name__ == "__main__":
# y_train.shape = (2806-y_test, nr_subjects) # y_train.shape = (2806-y_test, nr_subjects)
# y_test.shape = (y_test(from session nr. ?), nr_subjects) # y_test.shape = (y_test(from session nr. ?), nr_subjects)
#X_train, X_test, y_train, y_test = prepare_datasets_sessions(X, y, session_lengths, TEST_SESSION_NR) X_train, X_test, y_train, y_test = prepare_datasets_sessions(X, y, session_lengths, TEST_SESSION_NR)
#''' #'''
# ----- Make model ------ # ----- Make model ------
#model_GRU = GRU(input_shape=(1, 208)) # (timestep, 13*16 MFCC coefficients) #model_GRU = GRU(input_shape=(1, 208)) # (timestep, 13*16 MFCC coefficients)
#model_LSTM = LSTM(input_shape=(1, 208)) # (timestep, 13*16 MFCC coefficients) #model_LSTM = LSTM(input_shape=(1, 208)) # (timestep, 13*16 MFCC coefficients)
model_CNN_1D = CNN(input_shape=(208, 1)) # (timestep, 13*16 MFCC coefficients)
model_CNN_1D.summary()
#model_GRU.summary() #model_GRU.summary()
#model_LSTM.summary() #model_LSTM.summary()
@ -386,7 +381,7 @@ if __name__ == "__main__":
# ----- Train network ------ # ----- Train network ------
#history_GRU = train(model_GRU, X_train, y_train, verbose=VERBOSE, batch_size=BATCH_SIZE, epochs=EPOCHS) #history_GRU = train(model_GRU, X_train, y_train, verbose=VERBOSE, batch_size=BATCH_SIZE, epochs=EPOCHS)
#history_LSTM = train(model_LSTM, X_train, y_train, verbose=VERBOSE, batch_size=BATCH_SIZE, epochs=EPOCHS) #history_LSTM = train(model_LSTM, X_train, y_train, verbose=VERBOSE, batch_size=BATCH_SIZE, epochs=EPOCHS)
history_CNN_1D = train(model_CNN_1D, np.reshape(X_train, (X_train.shape[0], 208, 1)), y_train, verbose=VERBOSE, batch_size=BATCH_SIZE, epochs=EPOCHS)
# ----- Plot train accuracy/error ----- # ----- Plot train accuracy/error -----
#plot_train_history(history) #plot_train_history(history)
@ -397,9 +392,11 @@ if __name__ == "__main__":
#print('\nTest accuracy GRU:', test_acc, '\n') #print('\nTest accuracy GRU:', test_acc, '\n')
#test_loss, test_acc = model_LSTM.evaluate(X_test, y_test, verbose=VERBOSE) #test_loss, test_acc = model_LSTM.evaluate(X_test, y_test, verbose=VERBOSE)
#print('\nTest accuracy LSTM:', test_acc, '\n') #print('\nTest accuracy LSTM:', test_acc, '\n')
test_loss, test_acc = model_CNN_1D.evaluate(np.reshape(X_test, (X_test.shape[0], 208, 1)), y_test, verbose=VERBOSE)
print('\nTest accuracy CNN_1D:', test_acc, '\n')
#''' #'''
'''
# ----- Cross validation ------ # ----- Cross validation ------
#average_GRU = session_cross_validation('GRU', X, y, session_lengths, NR_SESSIONS, BATCH_SIZE, EPOCHS) #average_GRU = session_cross_validation('GRU', X, y, session_lengths, NR_SESSIONS, BATCH_SIZE, EPOCHS)
#verage_LSTM = session_cross_validation('LSTM', X, y, session_lengths, NR_SESSIONS, BATCH_SIZE, EPOCHS) #verage_LSTM = session_cross_validation('LSTM', X, y, session_lengths, NR_SESSIONS, BATCH_SIZE, EPOCHS)
@ -410,6 +407,7 @@ if __name__ == "__main__":
#print('Crossvalidated GRU:', average_GRU) #print('Crossvalidated GRU:', average_GRU)
#print('Crossvalidated LSTM:', average_LSTM) #print('Crossvalidated LSTM:', average_LSTM)
#print('Crossvalidated FFN:', average_FFN) #print('Crossvalidated FFN:', average_FFN)
print('Crossvalidated CNN:', average_CNN) print('Cross-validated CNN:', average_CNN)
print('\n') print('\n')
'''