import numpy
from keras.layers import *
from keras.layers.convolutional import _Conv
from keras import backend as K
from keras.engine.topology import Layer
from keras import initializers
class Conv3DHighway(_Conv):
def __init__(self, filters,
kernel_size,
strides=(1, 1, 1),
padding='valid',
data_format=None,
dilation_rate=(1, 1, 1),
activation=None,
use_bias=True,
kernel_initializer='glorot_uniform',
bias_initializer='zeros',
kernel_regularizer=None,
bias_regularizer=None,
activity_regularizer=None,
kernel_constraint=None,
bias_constraint=None,
**kwargs):
super(Conv3DHighway, self).__init__(
rank=3,
filters=filters,
kernel_size=kernel_size,
strides=strides,
padding=padding,
data_format=data_format,
dilation_rate=dilation_rate,
activation=activation,
use_bias=use_bias,
kernel_initializer=kernel_initializer,
bias_initializer=bias_initializer,
kernel_regularizer=kernel_regularizer,
bias_regularizer=bias_regularizer,
activity_regularizer=activity_regularizer,
kernel_constraint=kernel_constraint,
bias_constraint=bias_constraint,
**kwargs)
self.input_spec = InputSpec(ndim=5)
def get_config(self):
config = super(Conv3DHighway, self).get_config()
config.pop('rank')
return config
def build(self, input_shape):
super(Conv3DHighway, self).build(input_shape)
if self.data_format == 'channels_first':
channel_axis = 1
else:
channel_axis = -1
if input_shape[channel_axis] is None:
raise ValueError('The channel dimension of the inputs '
'should be defined. Found `None`.')
input_dim = input_shape[channel_axis]
kernel_shape = self.kernel_size + (input_dim, self.filters)
self.kernel_gate = self.add_weight(shape=kernel_shape,
initializer=self.kernel_initializer,
name='kernel_gate',
regularizer=self.kernel_regularizer,
constraint=self.kernel_constraint)
if self.use_bias:
self.bias_gate = self.add_weight(shape=(self.filters,),
initializer=self.bias_initializer,
name='bias_gate',
regularizer=self.bias_regularizer,
constraint=self.bias_constraint)
else:
self.bias_gate = None
def get_output_shape_for(self, input_shape):
super(Conv3DHighway, self).get_output_shape_for(input_shape)
def call(self, inputs, mask=None):
# compute channels_firste candidate hidden state
# Arguments
'''
x: Tensor or variable.
kernel: kernel tensor.
strides: strides tuple.
padding: string, `"same"` or `"valid"`.
data_format: string, `"channels_last"` or `"channels_first"`.
Whether to use Theano or TensorFlow/CNTK data format
for inputs/kernels/outputs.
dilation_rate: tuple of 3 integers.
'''
transform = K.conv3d(
inputs,
self.kernel,
strides=self.strides,
padding=self.padding,
data_format=self.data_format,
dilation_rate=self.dilation_rate)
if self.use_bias:
transform = K.bias_add(
transform,
self.bias,
data_format=self.data_format)
if self.activation is not None:
transform = self.activation(transform)
transform_gate = K.conv3d(
inputs,
self.kernel_gate,
strides=self.strides,
padding=self.padding,
data_format=self.data_format,
dilation_rate=self.dilation_rate)
if self.use_bias:
transform = K.bias_add(
transform,
self.bias_gate,
data_format=self.data_format)
transform_gate = K.sigmoid(transform_gate)
carry_gate = 1.0 - transform_gate
return transform * transform_gate + inputs * carry_gate
