{ }
View source on GitHub |
1D transposed convolution layer.
Inherits From: Layer
, Operation
tf.keras.layers.Conv1DTranspose(
filters,
kernel_size,
strides=1,
padding='valid',
data_format=None,
dilation_rate=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
)
The need for transposed convolutions generally arise from the desire to use a transformation going in the opposite direction of a normal convolution, i.e., from something that has the shape of the output of some convolution to something that has the shape of its input while maintaining a connectivity pattern that is compatible with said convolution.
Input shape:
- If
data_format="channels_last"
: A 3D tensor with shape:(batch_shape, steps, channels)
- If
data_format="channels_first"
: A 3D tensor with shape:(batch_shape, channels, steps)
Output shape:
- If
data_format="channels_last"
: A 3D tensor with shape:(batch_shape, new_steps, filters)
- If
data_format="channels_first"
: A 3D tensor with shape:(batch_shape, filters, new_steps)
Returns | |
---|---|
A 3D tensor representing
activation(conv1d_transpose(inputs, kernel) + bias) .
|
Raises | |
---|---|
ValueError
|
when both strides > 1 and dilation_rate > 1 .
|
References:
Example:
x = np.random.rand(4, 10, 128)
y = keras.layers.Conv1DTranspose(32, 3, 2, activation='relu')(x)
print(y.shape)
(4, 21, 32)
Methods
from_config
@classmethod
from_config( config )
Creates a layer from its config.
This method is the reverse of get_config
,
capable of instantiating the same layer from the config
dictionary. It does not handle layer connectivity
(handled by Network), nor weights (handled by set_weights
).
Args | |
---|---|
config
|
A Python dictionary, typically the output of get_config. |
Returns | |
---|---|
A layer instance. |
symbolic_call
symbolic_call(
*args, **kwargs
)