2D convolution with separable filters.
tf.keras.backend.depthwise_conv2d(
x, depthwise_kernel, strides=(1, 1), padding='valid', data_format=None,
dilation_rate=(1, 1)
)
Arguments |
x
|
input tensor
|
depthwise_kernel
|
convolution kernel for the depthwise convolution.
|
strides
|
strides tuple (length 2).
|
padding
|
string, "same" or "valid" .
|
data_format
|
string, "channels_last" or "channels_first" .
|
dilation_rate
|
tuple of integers,
dilation rates for the separable convolution.
|
Raises |
ValueError
|
if data_format is neither channels_last or
channels_first .
|