-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathDeepConvLSTM_1.0.py
211 lines (158 loc) · 8.29 KB
/
DeepConvLSTM_1.0.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
# coding: utf-8
#We would recommend to download the OPPORTUNITY zip file from the UCI repository and then use the script to generate the data file.
# In[2]:
#get_ipython().system(u'wget https://archive.ics.uci.edu/ml/machine-learning-databases/00226/OpportunityUCIDataset.zip')
# In[3]:
#get_ipython().system(u'python preprocess_data.py -h')
# In[1]:
#get_ipython().system(u'python preprocess_data.py -i data/OpportunityUCIDataset.zip -o oppChallenge_gestures.data')
#DeepConvLSTM is defined as a neural netowrk which combines convolutional and recurrent layers. The convolutional
#layers act as feature extractors and provide abstract representations of the input sensor data in feature
#maps. The recurrent layers model the temporal dynamics of the activation of the feature maps.
# In[3]:
import tensorflow as tf
import lasagne
import theano
import time
import numpy as np
import cPickle as cp #serializing and de-serializing a Python object structure
import theano.tensor as T
from sliding_window import sliding_window
# Hardcoded number of sensor channels employed in the OPPORTUNITY challenge
NB_SENSOR_CHANNELS = 113
# Hardcoded number of classes in the gesture recognition problem
NUM_CLASSES = 18
# Hardcoded length of the sliding window mechanism employed to segment the data
SLIDING_WINDOW_LENGTH = 24
# Length of the input sequence after convolutional operations
FINAL_SEQUENCE_LENGTH = 8
# Hardcoded step of the sliding window mechanism employed to segment the data
SLIDING_WINDOW_STEP = 12
# Batch Size
BATCH_SIZE = 100
# Number filters convolutional layers
NUM_FILTERS = 64
# Size filters convolutional layers
FILTER_SIZE = 5
# Number of unit in the long short-term recurrent layers
NUM_UNITS_LSTM = 128
#Load the OPPORTUNITY processed dataset. Sensor data is segmented using a sliding window of fixed length. The class associated with each segment corresponds to the gesture which has been observed during that interval. Given a sliding window of length T, we choose the class of the sequence as the label at t=T, or in other words, the label of last sample in the window.
# In[14]:
def load_dataset(filename):
f = file(filename, 'rb')
data = cp.load(f)
f.close()
X_train, y_train = data[0]
X_test, y_test = data[1]
print(" ..from file {}".format(filename))
print(" ..reading instances: train {0}, test {1}".format(X_train.shape, X_test.shape))
X_train = X_train.astype(np.float32)
X_test = X_test.astype(np.float32)
# The targets are casted to int8 for GPU compatibility.
y_train = y_train.astype(np.uint8)
y_test = y_test.astype(np.uint8)
return X_train, y_train, X_test, y_test
print("Loading data...")
X_train, y_train, X_test, y_test = load_dataset('data/oppChallenge_gestures.data')
assert NB_SENSOR_CHANNELS == X_train.shape[1]
def opp_sliding_window(data_x, data_y, ws, ss):
data_x = sliding_window(data_x,(ws,data_x.shape[1]),(ss,1))
data_y = np.asarray([[i[-1]] for i in sliding_window(data_y,ws,ss)])
return data_x.astype(np.float32), data_y.reshape(len(data_y)).astype(np.uint8)
# Sensor data is segmented using a sliding window mechanism
X_test, y_test = opp_sliding_window(X_test, y_test, SLIDING_WINDOW_LENGTH, SLIDING_WINDOW_STEP)
print(" ..after sliding window (testing): inputs {0}, targets {1}".format(X_test.shape, y_test.shape))
# Data is reshaped since the input of the network is a 4 dimension tensor
X_test = X_test.reshape((-1, SLIDING_WINDOW_LENGTH, NB_SENSOR_CHANNELS, 1))
# Sensor data is segmented using a sliding window mechanism
X_train, y_train = opp_sliding_window(X_train, y_train, SLIDING_WINDOW_LENGTH, SLIDING_WINDOW_STEP)
print(" ..after sliding window (testing): inputs {0}, targets {1}".format(X_test.shape, y_test.shape))
# Data is reshaped since the input of the network is a 4 dimension tensor
X_train = X_train.reshape((-1, SLIDING_WINDOW_LENGTH, NB_SENSOR_CHANNELS, 1))
def one_hot(label):
"""convert label from dense to one hot
argument:
label: ndarray dense label ,shape: [sample_num,1]
return:
one_hot_label: ndarray one hot, shape: [sample_num,n_class]
"""
label_num = len(label)
new_label = label.reshape(label_num) # shape : [sample_num]
# because max is 5, and we will create 6 columns
n_values = np.max(new_label) + 1
return np.eye(n_values)[np.array(new_label, dtype=np.int32)]
y_test=one_hot(y_test)
y_train=one_hot(y_train)
print("data is ready")
def init_weights(shape):
return tf.Variable(tf.random_normal(shape, stddev=0.01))
def model(X, w, w2, w3, w4, rnnW, rnnB, lstm_size):
l1a = tf.nn.relu(tf.nn.conv2d(X, w, # l1a shape=(?, 28, 28, 32)
strides=[1, 1, 1, 1], padding='VALID'))
# l1 = tf.nn.max_pool(l1a, ksize=[1, 2, 2, 1], # l1 shape=(?, 14, 14, 32)
# strides=[1, 2, 2, 1], padding='SAME')
# l1 = tf.nn.dropout(l1, p_keep_conv)
l2a = tf.nn.relu(tf.nn.conv2d(l1a, w2, # l2a shape=(?, 14, 14, 64)
strides=[1, 1, 1, 1], padding='VALID'))
# l2 = tf.nn.max_pool(l2a, ksize=[1, 2, 2, 1], # l2 shape=(?, 7, 7, 64)
# strides=[1, 2, 2, 1], padding='SAME')
# l2 = tf.nn.dropout(l2, p_keep_conv)
l3a = tf.nn.relu(tf.nn.conv2d(l2a, w3, # l3a shape=(?, 7, 7, 128)
strides=[1, 1, 1, 1], padding='VALID'))
# l3 = tf.nn.max_pool(l3a, ksize=[1, 2, 2, 1], # l3 shape=(?, 4, 4, 128)
# strides=[1, 2, 2, 1], padding='SAME')
# l3 = tf.reshape(l3, [-1, w4.get_shape().as_list()[0]]) # reshape to (?, 2048)
# l3 = tf.nn.dropout(l3, p_keep_conv)
l4a = tf.nn.relu(tf.nn.conv2d(l3a, w4, # l3a shape=(?, 7, 7, 128)
strides=[1, 1, 1, 1], padding='VALID'))
shuff = tf.transpose(l4a, [1, 0, 2, 3])
shp1 = tf.reshape(shuff, [-1, lstm_size])
X_split = tf.split(shp1, 8, 0) # split them to time_step_size (28 arrays)
lstm_cell = tf.contrib.rnn.BasicLSTMCell(lstm_size, forget_bias=1.0, state_is_tuple=True)
# Stack two LSTM layers, both layers has the same shape
lstm_layers = tf.contrib.rnn.MultiRNNCell([lstm_cell] * 2)
outputs, _states = tf.contrib.rnn.static_rnn(lstm_layers, X_split, dtype=tf.float32)
print("tf net end")
# Linear activation
# Get the last output
return tf.matmul(outputs[-1], rnnW) + rnnB
X = tf.placeholder("float", [None, 24, 113, 1])
Y = tf.placeholder("float", [None, 18])
lstm_size = 128
w = init_weights([5, 1, 1, 64]) # 3x3x1 conv, 32 outputs
w2 = init_weights([5, 1, 64, 64]) # 3x3x32 conv, 64 outputs
w3 = init_weights([5, 1, 64, 64]) # 3x3x32 conv, 128 outputs
w4 = init_weights([5, 1, 64, 64]) # FC 128 * 4 * 4 inputs, 625 outputs
rnnW = init_weights([lstm_size, 128])
rnnB = init_weights([128])
pre_Y = model(X, w, w2, w3, w4, rnnW, rnnB,lstm_size)
print ("get cnn output");
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=pre_Y, labels=Y))
train_op = tf.train.RMSPropOptimizer(0.001, 0.9).minimize(cost)
predict_op = tf.argmax(pre_Y, 1)
session_conf = tf.ConfigProto()
session_conf.gpu_options.allow_growth = True
print("net work done")
def iterate_minibatches(inputs, targets, batchsize, shuffle=False):
assert len(inputs) == len(targets)
if shuffle:
indices = np.arange(len(inputs))
np.random.shuffle(indices)
for start_idx in range(0, len(inputs) - batchsize + 1, batchsize):
if shuffle:
excerpt = indices[start_idx:start_idx + batchsize]
else:
excerpt = slice(start_idx, start_idx + batchsize)
yield inputs[excerpt], targets[excerpt]
# Launch the graph in a session
with tf.Session(config=session_conf) as sess:
# you need to initialize all variables
tf.global_variables_initializer().run()
for i in range(100):
for start, end in zip(range(0, len(X_train), BATCH_SIZE), range(BATCH_SIZE, len(X_train)+1, BATCH_SIZE)):
sess.run(train_op, feed_dict={X: X_train[start:end], Y: y_train[start:end]})
test_indices = np.arange(len(X_test)) # Get A Test Batch
np.random.shuffle(test_indices)
test_indices = test_indices[0: 100]
print(i, np.mean(np.argmax(y_test[test_indices], axis=1) ==
sess.run(predict_op, feed_dict={X: X_test[test_indices]})))