lunny/keras
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Add convlstm3D layer

Browse Source
This commit is contained in:
Francois Chollet 2023-05-17 16:23:59 -07:00
parent b781ef7e53
commit 7525d0e182
6 changed files with 244 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
keras_core/layers/__init__.py
View File

@ -107,6 +107,7 @@ from keras_core.layers.reshaping.zero_padding3d import ZeroPadding3D
from keras_core.layers.rnn.bidirectional import Bidirectional
from keras_core.layers.rnn.conv_lstm1d import ConvLSTM1D
from keras_core.layers.rnn.conv_lstm2d import ConvLSTM2D
from keras_core.layers.rnn.conv_lstm3d import ConvLSTM3D
from keras_core.layers.rnn.gru import GRU
from keras_core.layers.rnn.lstm import LSTM
from keras_core.layers.rnn.rnn import RNN

13
keras_core/layers/rnn/conv_lstm.py
View File

@ -438,6 +438,13 @@ class ConvLSTM(RNN):
recurrent_constraint: Constraint function applied to
the `recurrent_kernel` weights matrix.
bias_constraint: Constraint function applied to the bias vector.
dropout: Float between 0 and 1.
Fraction of the units to drop for
the linear transformation of the inputs.
recurrent_dropout: Float between 0 and 1.
Fraction of the units to drop for
the linear transformation of the recurrent state.
seed: Random seed for dropout.
return_sequences: Boolean. Whether to return the last output
in the output sequence, or the full sequence. (default False)
return_state: Boolean Whether to return the last state
@ -447,12 +454,6 @@ class ConvLSTM(RNN):
stateful: Boolean (default False). If True, the last state
for each sample at index i in a batch will be used as initial
state for the sample of index i in the following batch.
dropout: Float between 0 and 1.
Fraction of the units to drop for
the linear transformation of the inputs.
recurrent_dropout: Float between 0 and 1.
Fraction of the units to drop for
the linear transformation of the recurrent state.
"""
def __init__(

3
keras_core/layers/rnn/conv_lstm1d.py
View File

@ -59,6 +59,7 @@ class ConvLSTM1D(ConvLSTM):
linear transformation of the inputs.
recurrent_dropout: Float between 0 and 1. Fraction of the units to drop
for the linear transformation of the recurrent state.
seed: Random seed for dropout.
return_sequences: Boolean. Whether to return the last output
in the output sequence, or the full sequence. Default: `False`.
return_state: Boolean. Whether to return the last state in addition
@ -143,6 +144,7 @@ class ConvLSTM1D(ConvLSTM):
bias_constraint=None,
dropout=0.0,
recurrent_dropout=0.0,
seed=None,
return_sequences=False,
return_state=False,
go_backwards=False,
@ -177,5 +179,6 @@ class ConvLSTM1D(ConvLSTM):
stateful=stateful,
dropout=dropout,
recurrent_dropout=recurrent_dropout,
seed=seed,
**kwargs
)

5
keras_core/layers/rnn/conv_lstm2d.py
View File

@ -59,6 +59,7 @@ class ConvLSTM2D(ConvLSTM):
linear transformation of the inputs.
recurrent_dropout: Float between 0 and 1. Fraction of the units to drop
for the linear transformation of the recurrent state.
seed: Random seed for dropout.
return_sequences: Boolean. Whether to return the last output
in the output sequence, or the full sequence. Default: `False`.
return_state: Boolean. Whether to return the last state in addition
@ -110,7 +111,7 @@ class ConvLSTM2D(ConvLSTM):
- Else, 4D tensor with shape: `(samples, filters, new_rows, new_cols)` if
`data_format='channels_first'`
or shape: `(samples, new_rows, new_cols, filters)` if
`data_format='channels_last'`.
`data_format='channels_last'`.
References:
@ -143,6 +144,7 @@ class ConvLSTM2D(ConvLSTM):
bias_constraint=None,
dropout=0.0,
recurrent_dropout=0.0,
seed=None,
return_sequences=False,
return_state=False,
go_backwards=False,
@ -177,5 +179,6 @@ class ConvLSTM2D(ConvLSTM):
stateful=stateful,
dropout=dropout,
recurrent_dropout=recurrent_dropout,
seed=seed,
**kwargs
)

183
keras_core/layers/rnn/conv_lstm3d.py Normal file
View File

@ -0,0 +1,183 @@
from keras_core.api_export import keras_core_export
from keras_core.layers.rnn.conv_lstm import ConvLSTM
@keras_core_export("keras_core.layers.ConvLSTM3D")
class ConvLSTM3D(ConvLSTM):
"""3D Convolutional LSTM.
Similar to an LSTM layer, but the input transformations
and recurrent transformations are both convolutional.
Args:
filters: int, the dimension of the output space (the number of filters
in the convolution).
kernel_size: int or tuple/list of 3 integers, specifying the size of the
convolution window.
strides: int or tuple/list of 3 integers, specifying the stride length
of the convolution. `strides > 1` is incompatible with
`dilation_rate > 1`.
padding: string, `"valid"` or `"same"` (case-insensitive).
`"valid"` means no padding. `"same"` results in padding evenly to
the left/right or up/down of the input such that output has the same
height/width dimension as the input.
data_format: string, either `"channels_last"` or `"channels_first"`.
The ordering of the dimensions in the inputs. `"channels_last"`
corresponds to inputs with shape `(batch, steps, features)`
while `"channels_first"` corresponds to inputs with shape
`(batch, features, steps)`. It defaults to the `image_data_format`
value found in your Keras config file at `~/.keras/keras.json`.
If you never set it, then it will be `"channels_last"`.
dilation_rate: int or tuple/list of 3 integers, specifying the dilation
rate to use for dilated convolution.
activation: Activation function to use. By default hyperbolic tangent
activation function is applied (`tanh(x)`).
recurrent_activation: Activation function to use for the recurrent step.
use_bias: Boolean, whether the layer uses a bias vector.
kernel_initializer: Initializer for the `kernel` weights matrix,
used for the linear transformation of the inputs.
recurrent_initializer: Initializer for the `recurrent_kernel` weights
matrix, used for the linear transformation of the recurrent state.
bias_initializer: Initializer for the bias vector.
unit_forget_bias: Boolean. If `True`, add 1 to the bias of the forget
gate at initialization.
Use in combination with `bias_initializer="zeros"`.
This is recommended in [Jozefowicz et al., 2015](
http://www.jmlr.org/proceedings/papers/v37/jozefowicz15.pdf)
kernel_regularizer: Regularizer function applied to the `kernel` weights
matrix.
recurrent_regularizer: Regularizer function applied to the
`recurrent_kernel` weights matrix.
bias_regularizer: Regularizer function applied to the bias vector.
activity_regularizer: Regularizer function applied to.
kernel_constraint: Constraint function applied to the `kernel` weights
matrix.
recurrent_constraint: Constraint function applied to the
`recurrent_kernel` weights matrix.
bias_constraint: Constraint function applied to the bias vector.
dropout: Float between 0 and 1. Fraction of the units to drop for the
linear transformation of the inputs.
recurrent_dropout: Float between 0 and 1. Fraction of the units to drop
for the linear transformation of the recurrent state.
seed: Random seed for dropout.
return_sequences: Boolean. Whether to return the last output
in the output sequence, or the full sequence. Default: `False`.
return_state: Boolean. Whether to return the last state in addition
to the output. Default: `False`.
go_backwards: Boolean (default: `False`).
If `True`, process the input sequence backwards and return the
reversed sequence.
stateful: Boolean (default False). If `True`, the last state
for each sample at index i in a batch will be used as initial
state for the sample of index i in the following batch.
unroll: Boolean (default: `False`).
If `True`, the network will be unrolled,
else a symbolic loop will be used.
Unrolling can speed-up a RNN,
although it tends to be more memory-intensive.
Unrolling is only suitable for short sequences.
Call arguments:
inputs: A 6D tensor.
mask: Binary tensor of shape `(samples, timesteps)` indicating whether a
given timestep should be masked.
training: Python boolean indicating whether the layer should behave in
training mode or in inference mode.
This is only relevant if `dropout` or `recurrent_dropout` are set.
initial_state: List of initial state tensors to be passed to the first
call of the cell.
Input shape:
- If `data_format='channels_first'`:
5D tensor with shape: `(samples, time, channels, *spatial_dims)`
- If `data_format='channels_last'`:
5D tensor with shape: `(samples, time, *spatial_dims, channels)`
Output shape:
- If `return_state`: a list of tensors. The first tensor is the output.
The remaining tensors are the last states,
each 4D tensor with shape: `(samples, filters, *spatial_dims)` if
`data_format='channels_first'`
or shape: `(samples, *spatial_dims, filters)` if
`data_format='channels_last'`.
- If `return_sequences`: 5D tensor with shape: `(samples, timesteps,
filters, *spatial_dims)` if data_format='channels_first'
or shape: `(samples, timesteps, *spatial_dims, filters)` if
`data_format='channels_last'`.
- Else, 4D tensor with shape: `(samples, filters, *spatial_dims)` if
`data_format='channels_first'`
or shape: `(samples, *spatial_dims, filters)` if
`data_format='channels_last'`.
References:
- [Shi et al., 2015](http://arxiv.org/abs/1506.04214v1)
(the current implementation does not include the feedback loop on the
cells output).
"""
def __init__(
self,
filters,
kernel_size,
strides=1,
padding="valid",
data_format=None,
dilation_rate=1,
activation="tanh",
recurrent_activation="hard_sigmoid",
use_bias=True,
kernel_initializer="glorot_uniform",
recurrent_initializer="orthogonal",
bias_initializer="zeros",
unit_forget_bias=True,
kernel_regularizer=None,
recurrent_regularizer=None,
bias_regularizer=None,
activity_regularizer=None,
kernel_constraint=None,
recurrent_constraint=None,
bias_constraint=None,
dropout=0.0,
recurrent_dropout=0.0,
seed=None,
return_sequences=False,
return_state=False,
go_backwards=False,
stateful=False,
**kwargs
):
super().__init__(
rank=3,
filters=filters,
kernel_size=kernel_size,
strides=strides,
padding=padding,
data_format=data_format,
dilation_rate=dilation_rate,
activation=activation,
recurrent_activation=recurrent_activation,
use_bias=use_bias,
kernel_initializer=kernel_initializer,
recurrent_initializer=recurrent_initializer,
bias_initializer=bias_initializer,
unit_forget_bias=unit_forget_bias,
kernel_regularizer=kernel_regularizer,
recurrent_regularizer=recurrent_regularizer,
bias_regularizer=bias_regularizer,
activity_regularizer=activity_regularizer,
kernel_constraint=kernel_constraint,
recurrent_constraint=recurrent_constraint,
bias_constraint=bias_constraint,
return_sequences=return_sequences,
return_state=return_state,
go_backwards=go_backwards,
stateful=stateful,
dropout=dropout,
recurrent_dropout=recurrent_dropout,
seed=seed,
**kwargs
)

46
keras_core/layers/rnn/conv_lstm3d_test.py Normal file
View File

@ -0,0 +1,46 @@
from keras_core import layers
from keras_core import testing
class ConvLSTM1DTest(testing.TestCase):
def test_basics(self):
self.run_layer_test(
layers.ConvLSTM3D,
init_kwargs={"filters": 5, "kernel_size": 3, "padding": "same"},
input_shape=(3, 2, 4, 4, 4, 3),
expected_output_shape=(3, 4, 4, 4, 5),
expected_num_trainable_weights=3,
expected_num_non_trainable_weights=0,
supports_masking=True,
)
self.run_layer_test(
layers.ConvLSTM3D,
init_kwargs={
"filters": 5,
"kernel_size": 3,
"padding": "valid",
"recurrent_dropout": 0.5,
},
input_shape=(3, 2, 8, 8, 8, 3),
call_kwargs={"training": True},
expected_output_shape=(3, 6, 6, 6, 5),
expected_num_trainable_weights=3,
expected_num_non_trainable_weights=0,
supports_masking=True,
)
self.run_layer_test(
layers.ConvLSTM3D,
init_kwargs={
"filters": 5,
"kernel_size": 3,
"padding": "valid",
"return_sequences": True,
},
input_shape=(3, 2, 8, 8, 8, 3),
expected_output_shape=(3, 2, 6, 6, 6, 5),
expected_num_trainable_weights=3,
expected_num_non_trainable_weights=0,
supports_masking=True,
)
# TODO: correctness testing