-
Notifications
You must be signed in to change notification settings - Fork 69
/
CNN.py
198 lines (174 loc) · 9.1 KB
/
CNN.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
# VI network using THEANO, takes batches of state input
from NNobj import *
from theano_utils import *
class cnn(NNobj):
"Class for a convolutional neural network, inthe style of LeNet/Alexnet"
def __init__(self, model="CNN", im_size=[28, 28], dropout=False, devtype="cpu", grad_check=False, reg=0,
batchsize=128):
self.im_size = im_size # input image size
self.model = model
self.reg = reg # regularization (currently not implemented)
self.batchsize = batchsize # batch size for training
np.random.seed(0)
print(model)
# theano.config.blas.ldflags = "-L/usr/local/lib -lopenblas"
# X input : l=3 stacked images: obstacle map, goal map, current state map
self.X = T.ftensor4(name="X")
self.y = T.bvector("y") # output action
l = 3
filter_sizes = [[50, 3, 3],
[50, 3, 3],
[100, 3, 3],
[100, 3, 3],
[100, 3, 3]]
poolings = [2, 1, 2, 1, 1]
self.cnn_net = CNN(in_x=self.X, in_x_channels=l, imsize=self.im_size,
batchsize=self.batchsize, filter_sizes=filter_sizes,
poolings=poolings)
self.p_of_y = self.cnn_net.output
self.params = self.cnn_net.params
# Total 1910 parameters
self.cost = -T.mean(T.log(self.p_of_y)[T.arange(self.y.shape[0]),
self.y], dtype=theano.config.floatX)
self.y_pred = T.argmax(self.p_of_y, axis=1)
self.err = T.mean(T.neq(self.y_pred, self.y.flatten()), dtype=theano.config.floatX)
self.computeloss = theano.function(inputs=[self.X, self.y],
outputs=[self.err, self.cost])
self.y_out = theano.function(inputs=[self.X], outputs=[self.y_pred])
self.updates = []
self.train = []
def run_training(self, input, stepsize=0.01, epochs=10, output='None', batch_size=128, grad_check=True,
profile=False, data_fraction=1):
# run training from input matlab data file, and save test data prediction in output file
# load data from Matlab file, including
# im_data: flattened images
# value_data: flattened reward image
# state_data: flattened state images
# label_data: one-hot vector for action (state difference)
matlab_data = sio.loadmat(input)
im_data = matlab_data["im_data"]
im_data = (im_data - 1)/255 # obstacles = 1, free zone = 0
value_data = matlab_data["value_data"]
state1_data = matlab_data["state_x_data"]
state2_data = matlab_data["state_y_data"]
label_data = matlab_data["label_data"]
y_data = label_data.astype('int8')
x_im_data = im_data.astype(theano.config.floatX)
x_im_data = x_im_data.reshape(-1, 1, self.im_size[0], self.im_size[1])
x_val_data = value_data.astype(theano.config.floatX)
x_val_data = x_val_data.reshape(-1, 1, self.im_size[0], self.im_size[1])
x_state_data = np.zeros_like(x_im_data)
for i in x_state_data.shape[0]:
pos1 = state1_data[i]
pos2 = state2_data[i]
x_state_data[i, 0, pos1, pos2] = 1
x_data = np.append(x_im_data, x_val_data, axis=1)
x_data = np.append(x_data, x_state_data, axis=1)
all_training_samples = int(6/7.0*x_data.shape[0])
training_samples = int(data_fraction * all_training_samples)
x_train = x_data[0:training_samples]
y_train = y_data[0:training_samples]
x_test = x_data[all_training_samples:]
y_test = y_data[all_training_samples:]
y_test = y_test.flatten()
sortinds = np.random.permutation(training_samples)
x_train = x_train[sortinds]
y_train = y_train[sortinds]
y_train = y_train.flatten()
self.updates = rmsprop_updates_T(self.cost, self.params, stepsize=stepsize)
self.train = theano.function(inputs=[self.X, self.y], outputs=[], updates=self.updates)
print fmt_row(10, ["Epoch", "Train NLL", "Train Err", "Test NLL", "Test Err", "Epoch Time"])
for i_epoch in xrange(int(epochs)):
tstart = time.time()
# do training
for start in xrange(0, x_train.shape[0], batch_size):
end = start+batch_size
if end <= x_train.shape[0]:
self.train(x_train[start:end], y_train[start:end])
elapsed = time.time() - tstart
# compute losses
trainerr = 0.
trainloss = 0.
testerr = 0.
testloss = 0.
num = 0
for start in xrange(0, x_test.shape[0], batch_size):
end = start+batch_size
if end <= x_test.shape[0]:
num += 1
trainerr_, trainloss_ = self.computeloss(x_train[start:end], y_train[start:end])
testerr_, testloss_ = self.computeloss(x_test[start:end], y_test[start:end])
trainerr += trainerr_
trainloss += trainloss_
testerr += testerr_
testloss += testloss_
print fmt_row(10, [i_epoch, trainloss/num, trainerr/num, testloss/num, testerr/num, elapsed])
def predict(self, input):
# NN output for a single input, read from file
matlab_data = sio.loadmat(input)
im_data = matlab_data["im_data"]
im_data = (im_data - 1)/255 # obstacles = 1, free zone = 0
# state_data = matlab_data["state_data"]
state_data = matlab_data["state_xy_data"]
value_data = matlab_data["value_data"]
x_im_test = im_data.astype(theano.config.floatX)
x_im_test = x_im_test.reshape(-1, 1, self.im_size[0], self.im_size[1])
x_val_test = value_data.astype(theano.config.floatX)
x_val_test = x_val_test.reshape(-1, 1, self.im_size[0], self.im_size[1])
x_state_test = np.zeros_like(x_im_test)
x_state_test[0, 0, state_data[0, 0], state_data[0, 1]] = 1
x_test = np.append(x_im_test, x_val_test, axis=1)
x_test = np.append(x_test, x_state_test, axis=1)
out = self.y_out(x_test)
return out[0][0]
def load_weights(self, infile="weight_dump.pk"):
dump = pickle.load(open(infile, 'r'))
[n.set_value(p) for n, p in zip(self.params, dump)]
def save_weights(self, outfile="weight_dump.pk"):
pickle.dump([n.get_value() for n in self.params], open(outfile, 'w'))
class CNN(object):
"""CNN network"""
def __init__(self, in_x, in_x_channels, imsize, batchsize=128,
filter_sizes=[[50, 3, 3], [100, 3, 3]], poolings=[2, 2]):
"""
Allocate a CNN network with shared variable internal parameters.
:type in_x: theano.tensor.dtensor4
:param in_x: symbolic input image tensor, of shape [batchsize, in_x_channels, imsize[0], imsize[1]]
Typically : first channel is image, second is the reward prior, third is the current state image.
:type in_x_channels: int32
:param in_x_channels: number of input channels
:type imsize: tuple or list of length 2
:param imsize: (image height, image width)
:type batchsize: int32
:param batchsize: batch size
:type filter_sizes: int32 list of int32 3-tuples
:param filter_sizes: list of filter sizes for each layer, each a list of 3 integers:
num_filters,filter_width,filter_height
:type batchsize: int32 list
:param batchsize: list of pooling ratios after each layer (assumed symmetric)
"""
assert len(filter_sizes) == len(poolings)
n_conv_layers = len(filter_sizes)
self.params = []
# first conv layer
prev_layer = ConvLayer(in_x, filter_shape=[filter_sizes[0][0], in_x_channels, filter_sizes[0][1],
filter_sizes[0][2]],
image_shape=[batchsize, in_x_channels, imsize[0], imsize[1]],
poolsize=(poolings[0], poolings[0]))
self.params = self.params + prev_layer.params
# then the rest of the conv layers
for l in range(1, n_conv_layers):
new_layer = ConvLayer(prev_layer.output,
filter_shape=[filter_sizes[l][0], prev_layer.out_shape[1], filter_sizes[l][1],
filter_sizes[l][2]],
image_shape=prev_layer.out_shape,
poolsize=(poolings[l], poolings[l]))
self.params = self.params + new_layer.params
prev_layer = new_layer
# fully connected layer
final_conv_shape = new_layer.out_shape
flat_conv_out = new_layer.output.flatten(ndim=2)
flat_shape = [final_conv_shape[0], final_conv_shape[1]*final_conv_shape[2]*final_conv_shape[3]]
self.w_o = init_weights_T(flat_shape[1], 8)
self.output = T.nnet.softmax(T.dot(flat_conv_out, self.w_o))
self.params = self.params + [self.w_o]