# Copyright 2015 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""Locally-connected layers.
"""
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import numpy as np

from tensorflow.python.keras import activations
from tensorflow.python.keras import backend as K
from tensorflow.python.keras import constraints
from tensorflow.python.keras import initializers
from tensorflow.python.keras import regularizers
from tensorflow.python.keras.engine.base_layer import Layer
from tensorflow.python.keras.engine.input_spec import InputSpec
from tensorflow.python.keras.utils import conv_utils
from tensorflow.python.keras.utils import tf_utils
from tensorflow.python.util.tf_export import keras_export


@keras_export('keras.layers.LocallyConnected1D')
class LocallyConnected1D(Layer):
  """Locally-connected layer for 1D inputs.

  The `LocallyConnected1D` layer works similarly to
  the `Conv1D` layer, except that weights are unshared,
  that is, a different set of filters is applied at each different patch
  of the input.

  Example:
  ```python
      # apply a unshared weight convolution 1d of length 3 to a sequence with
      # 10 timesteps, with 64 output filters
      model = Sequential()
      model.add(LocallyConnected1D(64, 3, input_shape=(10, 32)))
      # now model.output_shape == (None, 8, 64)
      # add a new conv1d on top
      model.add(LocallyConnected1D(32, 3))
      # now model.output_shape == (None, 6, 32)
  ```

  Arguments:
      filters: Integer, the dimensionality of the output space
          (i.e. the number of output filters in the convolution).
      kernel_size: An integer or tuple/list of a single integer,
          specifying the length of the 1D convolution window.
      strides: An integer or tuple/list of a single integer,
          specifying the stride length of the convolution.
          Specifying any stride value != 1 is incompatible with specifying
          any `dilation_rate` value != 1.
      padding: Currently only supports `"valid"` (case-insensitive).
          `"same"` may be supported in the future.
      data_format: A string,
          one of `channels_last` (default) or `channels_first`.
          The ordering of the dimensions in the inputs.
          `channels_last` corresponds to inputs with shape
          `(batch, length, channels)` while `channels_first`
          corresponds to inputs with shape
          `(batch, channels, length)`.
          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".
      activation: Activation function to use.
          If you don't specify anything, no activation is applied
          (ie. "linear" activation: `a(x) = x`).
      use_bias: Boolean, whether the layer uses a bias vector.
      kernel_initializer: Initializer for the `kernel` weights matrix.
      bias_initializer: Initializer for the bias vector.
      kernel_regularizer: Regularizer function applied to
          the `kernel` weights matrix.
      bias_regularizer: Regularizer function applied to the bias vector.
      activity_regularizer: Regularizer function applied to
          the output of the layer (its "activation")..
      kernel_constraint: Constraint function applied to the kernel matrix.
      bias_constraint: Constraint function applied to the bias vector.
      implementation: implementation mode, either `1` or `2`.
          `1` loops over input spatial locations to perform the forward pass.
          It is memory-efficient but performs a lot of (small) ops.

          `2` stores layer weights in a dense but sparsely-populated 2D matrix
          and implements the forward pass as a single matrix-multiply. It uses
          a lot of RAM but performs few (large) ops.

          Depending on the inputs, layer parameters, hardware, and
          `tf.executing_eagerly()` one implementation can be dramatically faster
          (e.g. 50X) than another.

          It is recommended to benchmark both in the setting of interest to pick
          the most efficient one (in terms of speed and memory usage).

          Following scenarios could benefit from setting `implementation=2`:
              - eager execution;
              - inference;
              - running on CPU;
              - large amount of RAM available;
              - small models (few filters, small kernel);
              - using `padding=same` (only possible with `implementation=2`).

  Input shape:
      3D tensor with shape: `(batch_size, steps, input_dim)`

  Output shape:
      3D tensor with shape: `(batch_size, new_steps, filters)`
      `steps` value might have changed due to padding or strides.
  """

  def __init__(self,
               filters,
               kernel_size,
               strides=1,
               padding='valid',
               data_format=None,
               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,
               implementation=1,
               **kwargs):
    super(LocallyConnected1D, self).__init__(**kwargs)
    self.filters = filters
    self.kernel_size = conv_utils.normalize_tuple(kernel_size, 1, 'kernel_size')
    self.strides = conv_utils.normalize_tuple(strides, 1, 'strides')
    self.padding = conv_utils.normalize_padding(padding)
    if self.padding != 'valid' and implementation == 1:
      raise ValueError('Invalid border mode for LocallyConnected1D '
                       '(only "valid" is supported if implementation is 1): '
                       + padding)
    self.data_format = conv_utils.normalize_data_format(data_format)
    self.activation = activations.get(activation)
    self.use_bias = use_bias
    self.kernel_initializer = initializers.get(kernel_initializer)
    self.bias_initializer = initializers.get(bias_initializer)
    self.kernel_regularizer = regularizers.get(kernel_regularizer)
    self.bias_regularizer = regularizers.get(bias_regularizer)
    self.activity_regularizer = regularizers.get(activity_regularizer)
    self.kernel_constraint = constraints.get(kernel_constraint)
    self.bias_constraint = constraints.get(bias_constraint)
    self.implementation = implementation
    self.input_spec = InputSpec(ndim=3)

  @tf_utils.shape_type_conversion
  def build(self, input_shape):
    if self.data_format == 'channels_first':
      input_dim, input_length = input_shape[1], input_shape[2]
    else:
      input_dim, input_length = input_shape[2], input_shape[1]

    if input_dim is None:
      raise ValueError('Axis 2 of input should be fully-defined. '
                       'Found shape:', input_shape)
    self.output_length = conv_utils.conv_output_length(
        input_length, self.kernel_size[0], self.padding, self.strides[0])

    if self.implementation == 1:
      self.kernel_shape = (self.output_length, self.kernel_size[0] * input_dim,
                           self.filters)

      self.kernel = self.add_weight(
          shape=self.kernel_shape,
          initializer=self.kernel_initializer,
          name='kernel',
          regularizer=self.kernel_regularizer,
          constraint=self.kernel_constraint)

    elif self.implementation == 2:
      if self.data_format == 'channels_first':
        self.kernel_shape = (input_dim, input_length,
                             self.filters, self.output_length)
      else:
        self.kernel_shape = (input_length, input_dim,
                             self.output_length, self.filters)

      self.kernel = self.add_weight(shape=self.kernel_shape,
                                    initializer=self.kernel_initializer,
                                    name='kernel',
                                    regularizer=self.kernel_regularizer,
                                    constraint=self.kernel_constraint)

      self.kernel_mask = get_locallyconnected_mask(
          input_shape=(input_length,),
          kernel_shape=self.kernel_size,
          strides=self.strides,
          padding=self.padding,
          data_format=self.data_format
      )

    else:
      raise ValueError('Unrecognized implementation mode: %d.'
                       % self.implementation)

    if self.use_bias:
      self.bias = self.add_weight(
          shape=(self.output_length, self.filters),
          initializer=self.bias_initializer,
          name='bias',
          regularizer=self.bias_regularizer,
          constraint=self.bias_constraint)
    else:
      self.bias = None

    if self.data_format == 'channels_first':
      self.input_spec = InputSpec(ndim=3, axes={1: input_dim})
    else:
      self.input_spec = InputSpec(ndim=3, axes={-1: input_dim})
    self.built = True

  @tf_utils.shape_type_conversion
  def compute_output_shape(self, input_shape):
    if self.data_format == 'channels_first':
      input_length = input_shape[2]
    else:
      input_length = input_shape[1]

    length = conv_utils.conv_output_length(input_length, self.kernel_size[0],
                                           self.padding, self.strides[0])

    if self.data_format == 'channels_first':
      return (input_shape[0], self.filters, length)
    elif self.data_format == 'channels_last':
      return (input_shape[0], length, self.filters)

  def call(self, inputs):
    if self.implementation == 1:
      output = K.local_conv(inputs, self.kernel, self.kernel_size, self.strides,
                            (self.output_length,), self.data_format)

    elif self.implementation == 2:
      output = local_conv_matmul(inputs, self.kernel, self.kernel_mask,
                                 self.compute_output_shape(inputs.shape))

    else:
      raise ValueError('Unrecognized implementation mode: %d.'
                       % self.implementation)

    if self.use_bias:
      output = K.bias_add(output, self.bias, data_format=self.data_format)

    output = self.activation(output)
    return output

  def get_config(self):
    config = {
        'filters':
            self.filters,
        'kernel_size':
            self.kernel_size,
        'strides':
            self.strides,
        'padding':
            self.padding,
        'data_format':
            self.data_format,
        'activation':
            activations.serialize(self.activation),
        'use_bias':
            self.use_bias,
        'kernel_initializer':
            initializers.serialize(self.kernel_initializer),
        'bias_initializer':
            initializers.serialize(self.bias_initializer),
        'kernel_regularizer':
            regularizers.serialize(self.kernel_regularizer),
        'bias_regularizer':
            regularizers.serialize(self.bias_regularizer),
        'activity_regularizer':
            regularizers.serialize(self.activity_regularizer),
        'kernel_constraint':
            constraints.serialize(self.kernel_constraint),
        'bias_constraint':
            constraints.serialize(self.bias_constraint),
        'implementation':
            self.implementation
    }
    base_config = super(LocallyConnected1D, self).get_config()
    return dict(list(base_config.items()) + list(config.items()))


@keras_export('keras.layers.LocallyConnected2D')
class LocallyConnected2D(Layer):
  """Locally-connected layer for 2D inputs.

  The `LocallyConnected2D` layer works similarly
  to the `Conv2D` layer, except that weights are unshared,
  that is, a different set of filters is applied at each
  different patch of the input.

  Examples:
  ```python
      # apply a 3x3 unshared weights convolution with 64 output filters on a
      32x32 image
      # with `data_format="channels_last"`:
      model = Sequential()
      model.add(LocallyConnected2D(64, (3, 3), input_shape=(32, 32, 3)))
      # now model.output_shape == (None, 30, 30, 64)
      # notice that this layer will consume (30*30)*(3*3*3*64) + (30*30)*64
      parameters

      # add a 3x3 unshared weights convolution on top, with 32 output filters:
      model.add(LocallyConnected2D(32, (3, 3)))
      # now model.output_shape == (None, 28, 28, 32)
  ```

  Arguments:
      filters: Integer, the dimensionality of the output space
          (i.e. the number of output filters in the convolution).
      kernel_size: An integer or tuple/list of 2 integers, specifying the
          width and height of the 2D convolution window.
          Can be a single integer to specify the same value for
          all spatial dimensions.
      strides: An integer or tuple/list of 2 integers,
          specifying the strides of the convolution along the width and height.
          Can be a single integer to specify the same value for
          all spatial dimensions.
      padding: Currently only support `"valid"` (case-insensitive).
          `"same"` will be supported in future.
      data_format: A string,
          one of `channels_last` (default) or `channels_first`.
          The ordering of the dimensions in the inputs.
          `channels_last` corresponds to inputs with shape
          `(batch, height, width, channels)` while `channels_first`
          corresponds to inputs with shape
          `(batch, channels, height, width)`.
          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".
      activation: Activation function to use.
          If you don't specify anything, no activation is applied
          (ie. "linear" activation: `a(x) = x`).
      use_bias: Boolean, whether the layer uses a bias vector.
      kernel_initializer: Initializer for the `kernel` weights matrix.
      bias_initializer: Initializer for the bias vector.
      kernel_regularizer: Regularizer function applied to
          the `kernel` weights matrix.
      bias_regularizer: Regularizer function applied to the bias vector.
      activity_regularizer: Regularizer function applied to
          the output of the layer (its "activation").
      kernel_constraint: Constraint function applied to the kernel matrix.
      bias_constraint: Constraint function applied to the bias vector.
      implementation: implementation mode, either `1` or `2`.
          `1` loops over input spatial locations to perform the forward pass.
          It is memory-efficient but performs a lot of (small) ops.

          `2` stores layer weights in a dense but sparsely-populated 2D matrix
          and implements the forward pass as a single matrix-multiply. It uses
          a lot of RAM but performs few (large) ops.

          Depending on the inputs, layer parameters, hardware, and
          `tf.executing_eagerly()` one implementation can be dramatically faster
          (e.g. 50X) than another.

          It is recommended to benchmark both in the setting of interest to pick
          the most efficient one (in terms of speed and memory usage).

          Following scenarios could benefit from setting `implementation=2`:
              - eager execution;
              - inference;
              - running on CPU;
              - large amount of RAM available;
              - small models (few filters, small kernel);
              - using `padding=same` (only possible with `implementation=2`).

  Input shape:
      4D tensor with shape:
      `(samples, channels, rows, cols)` if data_format='channels_first'
      or 4D tensor with shape:
      `(samples, rows, cols, channels)` if data_format='channels_last'.

  Output shape:
      4D tensor with shape:
      `(samples, filters, new_rows, new_cols)` if data_format='channels_first'
      or 4D tensor with shape:
      `(samples, new_rows, new_cols, filters)` if data_format='channels_last'.
      `rows` and `cols` values might have changed due to padding.
  """

  def __init__(self,
               filters,
               kernel_size,
               strides=(1, 1),
               padding='valid',
               data_format=None,
               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,
               implementation=1,
               **kwargs):
    super(LocallyConnected2D, self).__init__(**kwargs)
    self.filters = filters
    self.kernel_size = conv_utils.normalize_tuple(kernel_size, 2, 'kernel_size')
    self.strides = conv_utils.normalize_tuple(strides, 2, 'strides')
    self.padding = conv_utils.normalize_padding(padding)
    if self.padding != 'valid' and implementation == 1:
      raise ValueError('Invalid border mode for LocallyConnected2D '
                       '(only "valid" is supported if implementation is 1): '
                       + padding)
    self.data_format = conv_utils.normalize_data_format(data_format)
    self.activation = activations.get(activation)
    self.use_bias = use_bias
    self.kernel_initializer = initializers.get(kernel_initializer)
    self.bias_initializer = initializers.get(bias_initializer)
    self.kernel_regularizer = regularizers.get(kernel_regularizer)
    self.bias_regularizer = regularizers.get(bias_regularizer)
    self.activity_regularizer = regularizers.get(activity_regularizer)
    self.kernel_constraint = constraints.get(kernel_constraint)
    self.bias_constraint = constraints.get(bias_constraint)
    self.implementation = implementation
    self.input_spec = InputSpec(ndim=4)

  @tf_utils.shape_type_conversion
  def build(self, input_shape):
    if self.data_format == 'channels_last':
      input_row, input_col = input_shape[1:-1]
      input_filter = input_shape[3]
    else:
      input_row, input_col = input_shape[2:]
      input_filter = input_shape[1]
    if input_row is None or input_col is None:
      raise ValueError('The spatial dimensions of the inputs to '
                       ' a LocallyConnected2D layer '
                       'should be fully-defined, but layer received '
                       'the inputs shape ' + str(input_shape))
    output_row = conv_utils.conv_output_length(input_row, self.kernel_size[0],
                                               self.padding, self.strides[0])
    output_col = conv_utils.conv_output_length(input_col, self.kernel_size[1],
                                               self.padding, self.strides[1])
    self.output_row = output_row
    self.output_col = output_col

    if self.implementation == 1:
      self.kernel_shape = (
          output_row * output_col,
          self.kernel_size[0] * self.kernel_size[1] * input_filter,
          self.filters)

      self.kernel = self.add_weight(
          shape=self.kernel_shape,
          initializer=self.kernel_initializer,
          name='kernel',
          regularizer=self.kernel_regularizer,
          constraint=self.kernel_constraint)

    elif self.implementation == 2:
      if self.data_format == 'channels_first':
        self.kernel_shape = (input_filter, input_row, input_col,
                             self.filters, self.output_row, self.output_col)
      else:
        self.kernel_shape = (input_row, input_col, input_filter,
                             self.output_row, self.output_col, self.filters)

      self.kernel = self.add_weight(shape=self.kernel_shape,
                                    initializer=self.kernel_initializer,
                                    name='kernel',
                                    regularizer=self.kernel_regularizer,
                                    constraint=self.kernel_constraint)

      self.kernel_mask = get_locallyconnected_mask(
          input_shape=(input_row, input_col),
          kernel_shape=self.kernel_size,
          strides=self.strides,
          padding=self.padding,
          data_format=self.data_format
      )

    else:
      raise ValueError('Unrecognized implementation mode: %d.'
                       % self.implementation)

    if self.use_bias:
      self.bias = self.add_weight(
          shape=(output_row, output_col, self.filters),
          initializer=self.bias_initializer,
          name='bias',
          regularizer=self.bias_regularizer,
          constraint=self.bias_constraint)
    else:
      self.bias = None
    if self.data_format == 'channels_first':
      self.input_spec = InputSpec(ndim=4, axes={1: input_filter})
    else:
      self.input_spec = InputSpec(ndim=4, axes={-1: input_filter})
    self.built = True

  @tf_utils.shape_type_conversion
  def compute_output_shape(self, input_shape):
    if self.data_format == 'channels_first':
      rows = input_shape[2]
      cols = input_shape[3]
    elif self.data_format == 'channels_last':
      rows = input_shape[1]
      cols = input_shape[2]

    rows = conv_utils.conv_output_length(rows, self.kernel_size[0],
                                         self.padding, self.strides[0])
    cols = conv_utils.conv_output_length(cols, self.kernel_size[1],
                                         self.padding, self.strides[1])

    if self.data_format == 'channels_first':
      return (input_shape[0], self.filters, rows, cols)
    elif self.data_format == 'channels_last':
      return (input_shape[0], rows, cols, self.filters)

  def call(self, inputs):
    if self.implementation == 1:
      output = K.local_conv(inputs, self.kernel, self.kernel_size, self.strides,
                            (self.output_row, self.output_col),
                            self.data_format)

    elif self.implementation == 2:
      output = local_conv_matmul(inputs, self.kernel, self.kernel_mask,
                                 self.compute_output_shape(inputs.shape))

    else:
      raise ValueError('Unrecognized implementation mode: %d.'
                       % self.implementation)

    if self.use_bias:
      output = K.bias_add(output, self.bias, data_format=self.data_format)

    output = self.activation(output)
    return output

  def get_config(self):
    config = {
        'filters':
            self.filters,
        'kernel_size':
            self.kernel_size,
        'strides':
            self.strides,
        'padding':
            self.padding,
        'data_format':
            self.data_format,
        'activation':
            activations.serialize(self.activation),
        'use_bias':
            self.use_bias,
        'kernel_initializer':
            initializers.serialize(self.kernel_initializer),
        'bias_initializer':
            initializers.serialize(self.bias_initializer),
        'kernel_regularizer':
            regularizers.serialize(self.kernel_regularizer),
        'bias_regularizer':
            regularizers.serialize(self.bias_regularizer),
        'activity_regularizer':
            regularizers.serialize(self.activity_regularizer),
        'kernel_constraint':
            constraints.serialize(self.kernel_constraint),
        'bias_constraint':
            constraints.serialize(self.bias_constraint),
        'implementation':
            self.implementation
    }
    base_config = super(LocallyConnected2D, self).get_config()
    return dict(list(base_config.items()) + list(config.items()))


def get_locallyconnected_mask(input_shape,
                              kernel_shape,
                              strides,
                              padding,
                              data_format):
  """Return a mask representing connectivity of a locally-connected operation.

  This method returns a masking numpy array of 0s and 1s (of type `np.float32`)
  that, when element-wise multiplied with a fully-connected weight tensor, masks
  out the weights between disconnected input-output pairs and thus implements
  local connectivity through a sparse fully-connected weight tensor.

  Assume an unshared convolution with given parameters is applied to an input
  having N spatial dimensions with `input_shape = (d_in1, ..., d_inN)`
  to produce an output with spatial shape `(d_out1, ..., d_outN)` (determined
  by layer parameters such as `strides`).

  This method returns a mask which can be broadcast-multiplied (element-wise)
  with a 2*(N+1)-D weight matrix (equivalent to a fully-connected layer between
  (N+1)-D activations (N spatial + 1 channel dimensions for input and output)
  to make it perform an unshared convolution with given `kernel_shape`,
  `strides`, `padding` and `data_format`.

  Arguments:
    input_shape: tuple of size N: `(d_in1, ..., d_inN)`
                 spatial shape of the input.
    kernel_shape: tuple of size N, spatial shape of the convolutional kernel
                  / receptive field.
    strides: tuple of size N, strides along each spatial dimension.
    padding: type of padding, string `"same"` or `"valid"`.
    data_format: a string, `"channels_first"` or `"channels_last"`.

  Returns:
    a `np.float32`-type `np.ndarray` of shape
    `(1, d_in1, ..., d_inN, 1, d_out1, ..., d_outN)`
    if `data_format == `"channels_first"`, or
    `(d_in1, ..., d_inN, 1, d_out1, ..., d_outN, 1)`
    if `data_format == "channels_last"`.

  Raises:
    ValueError: if `data_format` is neither `"channels_first"` nor
                `"channels_last"`.
  """
  mask = conv_utils.conv_kernel_mask(
      input_shape=input_shape,
      kernel_shape=kernel_shape,
      strides=strides,
      padding=padding
  )

  ndims = int(mask.ndim / 2)

  if data_format == 'channels_first':
    mask = np.expand_dims(mask, 0)
    mask = np.expand_dims(mask, -ndims - 1)

  elif data_format == 'channels_last':
    mask = np.expand_dims(mask, ndims)
    mask = np.expand_dims(mask, -1)

  else:
    raise ValueError('Unrecognized data_format: ' + str(data_format))

  return mask


def local_conv_matmul(inputs, kernel, kernel_mask, output_shape):
  """Apply N-D convolution with un-shared weights using a single matmul call.

  This method outputs `inputs . (kernel * kernel_mask)`
  (with `.` standing for matrix-multiply and `*` for element-wise multiply)
  and requires a precomputed `kernel_mask` to zero-out weights in `kernel` and
  hence perform the same operation as a convolution with un-shared
  (the remaining entries in `kernel`) weights. It also does the necessary
  reshapes to make `inputs` and `kernel` 2-D and `output` (N+2)-D.

  Arguments:
      inputs: (N+2)-D tensor with shape
          `(batch_size, channels_in, d_in1, ..., d_inN)`
          or
          `(batch_size, d_in1, ..., d_inN, channels_in)`.
      kernel: the unshared weights for N-D convolution,
          an (N+2)-D tensor of shape:
          `(d_in1, ..., d_inN, channels_in, d_out2, ..., d_outN, channels_out)`
          or
          `(channels_in, d_in1, ..., d_inN, channels_out, d_out2, ..., d_outN)`,
          with the ordering of channels and spatial dimensions matching
          that of the input.
          Each entry is the weight between a particular input and
          output location, similarly to a fully-connected weight matrix.
      kernel_mask: a float 0/1 mask tensor of shape:
           `(d_in1, ..., d_inN, 1, d_out2, ..., d_outN, 1)`
           or
           `(1, d_in1, ..., d_inN, 1, d_out2, ..., d_outN)`,
           with the ordering of singleton and spatial dimensions
           matching that of the input.
           Mask represents the connectivity pattern of the layer and is
           precomputed elsewhere based on layer parameters: stride,
           padding, and the receptive field shape.
      output_shape: a tuple of (N+2) elements representing the output shape:
          `(batch_size, channels_out, d_out1, ..., d_outN)`
          or
          `(batch_size, d_out1, ..., d_outN, channels_out)`,
          with the ordering of channels and spatial dimensions matching that of
          the input.

  Returns:
      Output (N+2)-D tensor with shape `output_shape`.
  """
  inputs_flat = K.reshape(inputs, (K.shape(inputs)[0], -1))

  kernel = kernel_mask * kernel
  kernel = make_2d(kernel, split_dim=K.ndim(kernel) // 2)

  output_flat = K.math_ops.sparse_matmul(inputs_flat, kernel, b_is_sparse=True)
  output = K.reshape(output_flat,
                     [K.shape(output_flat)[0],] + output_shape.as_list()[1:])
  return output


def make_2d(tensor, split_dim):
  """Reshapes an N-dimensional tensor into a 2D tensor.

  Dimensions before (excluding) and after (including) `split_dim` are grouped
  together.

  Arguments:
    tensor: a tensor of shape `(d0, ..., d(N-1))`.
    split_dim: an integer from 1 to N-1, index of the dimension to group
        dimensions before (excluding) and after (including).

  Returns:
    Tensor of shape
    `(d0 * ... * d(split_dim-1), d(split_dim) * ... * d(N-1))`.
  """
  shape = K.array_ops.shape(tensor)
  in_dims = shape[:split_dim]
  out_dims = shape[split_dim:]

  in_size = K.math_ops.reduce_prod(in_dims)
  out_size = K.math_ops.reduce_prod(out_dims)

  return K.array_ops.reshape(tensor, (in_size, out_size))