def get_output(self, train=False):
X = self.get_input(train)
newshape = (X.shape[0]*X.shape[1], X.shape[2], X.shape[3], X.shape[4])
Y = theano.tensor.reshape(X, newshape) #collapse num_samples and num_timesteps
border_mode = self.border_mode
if on_gpu() and dnn.dnn_available():
if border_mode == 'same':
assert(self.subsample == (1, 1))
pad_x = (self.nb_row - self.subsample[0]) // 2
pad_y = (self.nb_col - self.subsample[1]) // 2
conv_out = dnn.dnn_conv(img=Y,
kerns=self.W,
border_mode=(pad_x, pad_y))
else:
conv_out = dnn.dnn_conv(img=Y,
kerns=self.W,
border_mode=border_mode,
subsample=self.subsample)
else:
if border_mode == 'same':
border_mode = 'full'
conv_out = theano.tensor.nnet.conv.conv2d(Y, self.W,
border_mode=border_mode, subsample=self.subsample)
if self.border_mode == 'same':
shift_x = (self.nb_row - 1) // 2
shift_y = (self.nb_col - 1) // 2
conv_out = conv_out[:, :, shift_x:Y.shape[2] + shift_x, shift_y:Y.shape[3] + shift_y]
output = self.activation(conv_out + self.b.dimshuffle('x', 0, 'x', 'x'))
newshape = (X.shape[0], X.shape[1], output.shape[1], output.shape[2], output.shape[3])
return theano.tensor.reshape(output, newshape)
def get_output(self, train):
X = self.get_input(train)
border_mode = self.border_mode
if on_gpu() and dnn.dnn_available():
if border_mode == 'same':
assert (self.subsample == (1, 1))
pad_x = (self.nb_row - self.subsample[0]) // 2
pad_y = (self.nb_col - self.subsample[1]) // 2
conv_out = dnn.dnn_conv(img=X,
kerns=self.W,
border_mode=(pad_x, pad_y))
else:
conv_out = dnn.dnn_conv(img=X,
kerns=self.W,
border_mode=border_mode,
subsample=self.subsample)
else:
if border_mode == 'same':
border_mode = 'full'
conv_out = T.nnet.conv.conv2d(X,
self.W,
border_mode=border_mode,
subsample=self.subsample)
if self.border_mode == 'same':
shift_x = (self.nb_row - 1) // 2
shift_y = (self.nb_col - 1) // 2
conv_out = conv_out[:, :, shift_x:X.shape[2] + shift_x,
shift_y:X.shape[3] + shift_y]
return self.activation(conv_out)
def get_output(self, train=False):
X = self.get_input(train)
newshape = (X.shape[0] * X.shape[1], X.shape[2], X.shape[3],
X.shape[4])
Y = theano.tensor.reshape(
X, newshape) #collapse num_samples and num_timesteps
border_mode = self.border_mode
if on_gpu() and dnn.dnn_available():
if border_mode == 'same':
assert (self.subsample == (1, 1))
pad_x = (self.nb_row - self.subsample[0]) // 2
pad_y = (self.nb_col - self.subsample[1]) // 2
conv_out = dnn.dnn_conv(img=Y,
kerns=self.W,
border_mode=(pad_x, pad_y))
else:
conv_out = dnn.dnn_conv(img=Y,
kerns=self.W,
border_mode=border_mode,
subsample=self.subsample)
else:
if border_mode == 'same':
border_mode = 'full'
conv_out = theano.tensor.nnet.conv.conv2d(Y,
self.W,
border_mode=border_mode,
subsample=self.subsample)
if self.border_mode == 'same':
shift_x = (self.nb_row - 1) // 2
shift_y = (self.nb_col - 1) // 2
conv_out = conv_out[:, :, shift_x:Y.shape[2] + shift_x,
shift_y:Y.shape[3] + shift_y]
output = self.activation(conv_out +
self.b.dimshuffle('x', 0, 'x', 'x'))
newshape = (X.shape[0], X.shape[1], output.shape[1], output.shape[2],
output.shape[3])
return theano.tensor.reshape(output, newshape)
def get_output(self, train):
X = self.get_input(train)
border_mode = self.border_mode
if on_gpu() and dnn.dnn_available():
if border_mode == "same":
assert self.subsample == (1, 1)
pad_x = (self.nb_row - self.subsample[0]) // 2
pad_y = (self.nb_col - self.subsample[1]) // 2
conv_out = dnn.dnn_conv(img=X, kerns=self.W, border_mode=(pad_x, pad_y))
else:
conv_out = dnn.dnn_conv(img=X, kerns=self.W, border_mode=border_mode, subsample=self.subsample)
else:
if border_mode == "same":
border_mode = "full"
conv_out = T.nnet.conv.conv2d(X, self.W, border_mode=border_mode, subsample=self.subsample)
if self.border_mode == "same":
shift_x = (self.nb_row - 1) // 2
shift_y = (self.nb_col - 1) // 2
conv_out = conv_out[:, :, shift_x : X.shape[2] + shift_x, shift_y : X.shape[3] + shift_y]
return self.activation(conv_out)
from keras.layers.core import Layer
import theano.tensor as T
from theano.tensor.signal import downsample
from keras import activations, initializations, regularizers, constraints
from keras.utils.theano_utils import shared_zeros, floatX, on_gpu
from keras.utils.generic_utils import make_tuple
from keras.regularizers import ActivityRegularizer, Regularizer
if on_gpu():
from theano.sandbox.cuda import dnn
class Convolution2DNoBias(Layer):
def __init__(
self,
nb_filter,
stack_size,
nb_row,
nb_col,
init="glorot_uniform",
activation="linear",
weights=None,
border_mode="valid",
subsample=(1, 1),
W_regularizer=None,
b_regularizer=None,
activity_regularizer=None,
W_constraint=None,
b_constraint=None,
):
print '[new keras detected]'
from keras.backend import zeros as shared_zeros
def on_gpu():
return theano.config.device[:3] == 'gpu'
else:
from keras.utils.theano_utils import shared_zeros, floatX, on_gpu
from keras.utils.generic_utils import make_tuple
from keras.regularizers import ActivityRegularizer, Regularizer
from six.moves import zip
from keras.layers.core import Layer
if on_gpu():
from theano.sandbox.cuda import dnn
def conv_output_length(input_length, filter_size, border_mode, stride):
if input_length is None:
return None
assert border_mode in {'same', 'full', 'valid'}
if border_mode == 'same':
output_length = input_length
elif border_mode == 'full':
output_length = input_length + filter_size - 1
elif border_mode == 'valid':
output_length = input_length - filter_size + 1
return (output_length + stride - 1) // stride
