# Module: tf.keras.layers
|---| | [![](https://www.tensorflow.org/images/tf_logo_32px.png) TensorFlow 2 version](https://www.tensorflow.org/api_docs/python/tf/keras/layers) | Keras layers API. #### View aliases **Compat aliases for migration** See [Migration guide](https://www.tensorflow.org/guide/migrate) for more details. [`tf.compat.v1.keras.layers`](https://www.tensorflow.org/api_docs/python/tf/compat/v1/keras/layers)
## Classes [`class AbstractRNNCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/AbstractRNNCell): Abstract object representing an RNN cell. [`class Activation`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Activation): Applies an activation function to an output. [`class ActivityRegularization`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/ActivityRegularization): Layer that applies an update to the cost function based input activity. [`class Add`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Add): Layer that adds a list of inputs. [`class AdditiveAttention`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/AdditiveAttention): Additive attention layer, a.k.a. Bahdanau-style attention. [`class AlphaDropout`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/AlphaDropout): Applies Alpha Dropout to the input. [`class Attention`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Attention): Dot-product attention layer, a.k.a. Luong-style attention. [`class Average`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Average): Layer that averages a list of inputs. [`class AveragePooling1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/AveragePooling1D): Average pooling for temporal data. [`class AveragePooling2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/AveragePooling2D): Average pooling operation for spatial data. [`class AveragePooling3D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/AveragePooling3D): Average pooling operation for 3D data (spatial or spatio-temporal). [`class AvgPool1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/AveragePooling1D): Average pooling for temporal data. [`class AvgPool2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/AveragePooling2D): Average pooling operation for spatial data. [`class AvgPool3D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/AveragePooling3D): Average pooling operation for 3D data (spatial or spatio-temporal). [`class BatchNormalization`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/BatchNormalization): Base class of Batch normalization layer (Ioffe and Szegedy, 2014). [`class Bidirectional`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Bidirectional): Bidirectional wrapper for RNNs. [`class Concatenate`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Concatenate): Layer that concatenates a list of inputs. [`class Conv1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Conv1D): 1D convolution layer (e.g. temporal convolution). [`class Conv2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Conv2D): 2D convolution layer (e.g. spatial convolution over images). [`class Conv2DTranspose`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Conv2DTranspose): Transposed convolution layer (sometimes called Deconvolution). [`class Conv3D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Conv3D): 3D convolution layer (e.g. spatial convolution over volumes). [`class Conv3DTranspose`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Conv3DTranspose): Transposed convolution layer (sometimes called Deconvolution). [`class ConvLSTM2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/ConvLSTM2D): Convolutional LSTM. [`class Convolution1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Conv1D): 1D convolution layer (e.g. temporal convolution). [`class Convolution2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Conv2D): 2D convolution layer (e.g. spatial convolution over images). [`class Convolution2DTranspose`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Conv2DTranspose): Transposed convolution layer (sometimes called Deconvolution). [`class Convolution3D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Conv3D): 3D convolution layer (e.g. spatial convolution over volumes). [`class Convolution3DTranspose`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Conv3DTranspose): Transposed convolution layer (sometimes called Deconvolution). [`class Cropping1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Cropping1D): Cropping layer for 1D input (e.g. temporal sequence). [`class Cropping2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Cropping2D): Cropping layer for 2D input (e.g. picture). [`class Cropping3D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Cropping3D): Cropping layer for 3D data (e.g. spatial or spatio-temporal). [`class CuDNNGRU`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/CuDNNGRU): Fast GRU implementation backed by cuDNN. [`class CuDNNLSTM`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/CuDNNLSTM): Fast LSTM implementation backed by cuDNN. [`class Dense`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Dense): Just your regular densely-connected NN layer. [`class DenseFeatures`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/DenseFeatures): A layer that produces a dense `Tensor` based on given `feature_columns`. [`class DepthwiseConv2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/DepthwiseConv2D): Depthwise separable 2D convolution. [`class Dot`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Dot): Layer that computes a dot product between samples in two tensors. [`class Dropout`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Dropout): Applies Dropout to the input. [`class ELU`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/ELU): Exponential Linear Unit. [`class Embedding`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Embedding): Turns positive integers (indexes) into dense vectors of fixed size. [`class Flatten`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Flatten): Flattens the input. Does not affect the batch size. [`class GRU`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/GRU): Gated Recurrent Unit - Cho et al. 2014. [`class GRUCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/GRUCell): Cell class for the GRU layer. [`class GaussianDropout`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/GaussianDropout): Apply multiplicative 1-centered Gaussian noise. [`class GaussianNoise`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/GaussianNoise): Apply additive zero-centered Gaussian noise. [`class GlobalAveragePooling1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/GlobalAveragePooling1D): Global average pooling operation for temporal data. [`class GlobalAveragePooling2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/GlobalAveragePooling2D): Global average pooling operation for spatial data. [`class GlobalAveragePooling3D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/GlobalAveragePooling3D): Global Average pooling operation for 3D data. [`class GlobalAvgPool1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/GlobalAveragePooling1D): Global average pooling operation for temporal data. [`class GlobalAvgPool2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/GlobalAveragePooling2D): Global average pooling operation for spatial data. [`class GlobalAvgPool3D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/GlobalAveragePooling3D): Global Average pooling operation for 3D data. [`class GlobalMaxPool1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/GlobalMaxPool1D): Global max pooling operation for temporal data. [`class GlobalMaxPool2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/GlobalMaxPool2D): Global max pooling operation for spatial data. [`class GlobalMaxPool3D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/GlobalMaxPool3D): Global Max pooling operation for 3D data. [`class GlobalMaxPooling1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/GlobalMaxPool1D): Global max pooling operation for temporal data. [`class GlobalMaxPooling2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/GlobalMaxPool2D): Global max pooling operation for spatial data. [`class GlobalMaxPooling3D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/GlobalMaxPool3D): Global Max pooling operation for 3D data. [`class InputLayer`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/InputLayer): Layer to be used as an entry point into a Network (a graph of layers). [`class InputSpec`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/layers/InputSpec): Specifies the ndim, dtype and shape of every input to a layer. [`class LSTM`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/LSTM): Long Short-Term Memory layer - Hochreiter 1997. [`class LSTMCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/LSTMCell): Cell class for the LSTM layer. [`class Lambda`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Lambda): Wraps arbitrary expressions as a `Layer` object. [`class Layer`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Layer): Base layer class. [`class LayerNormalization`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/LayerNormalization): Layer normalization layer (Ba et al., 2016). [`class LeakyReLU`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/LeakyReLU): Leaky version of a Rectified Linear Unit. [`class LocallyConnected1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/LocallyConnected1D): Locally-connected layer for 1D inputs. [`class LocallyConnected2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/LocallyConnected2D): Locally-connected layer for 2D inputs. [`class Masking`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Masking): Masks a sequence by using a mask value to skip timesteps. [`class MaxPool1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/MaxPool1D): Max pooling operation for temporal data. [`class MaxPool2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/MaxPool2D): Max pooling operation for spatial data. [`class MaxPool3D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/MaxPool3D): Max pooling operation for 3D data (spatial or spatio-temporal). [`class MaxPooling1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/MaxPool1D): Max pooling operation for temporal data. [`class MaxPooling2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/MaxPool2D): Max pooling operation for spatial data. [`class MaxPooling3D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/MaxPool3D): Max pooling operation for 3D data (spatial or spatio-temporal). [`class Maximum`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Maximum): Layer that computes the maximum (element-wise) a list of inputs. [`class Minimum`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Minimum): Layer that computes the minimum (element-wise) a list of inputs. [`class Multiply`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Multiply): Layer that multiplies (element-wise) a list of inputs. [`class PReLU`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/PReLU): Parametric Rectified Linear Unit. [`class Permute`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Permute): Permutes the dimensions of the input according to a given pattern. [`class RNN`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/RNN): Base class for recurrent layers. [`class ReLU`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/ReLU): Rectified Linear Unit activation function. [`class RepeatVector`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/RepeatVector): Repeats the input n times. [`class Reshape`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Reshape): Reshapes an output to a certain shape. [`class SeparableConv1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/SeparableConv1D): Depthwise separable 1D convolution. [`class SeparableConv2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/SeparableConv2D): Depthwise separable 2D convolution. [`class SeparableConvolution1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/SeparableConv1D): Depthwise separable 1D convolution. [`class SeparableConvolution2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/SeparableConv2D): Depthwise separable 2D convolution. [`class SimpleRNN`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/SimpleRNN): Fully-connected RNN where the output is to be fed back to input. [`class SimpleRNNCell`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/SimpleRNNCell): Cell class for SimpleRNN. [`class Softmax`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Softmax): Softmax activation function. [`class SpatialDropout1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/SpatialDropout1D): Spatial 1D version of Dropout. [`class SpatialDropout2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/SpatialDropout2D): Spatial 2D version of Dropout. [`class SpatialDropout3D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/SpatialDropout3D): Spatial 3D version of Dropout. [`class StackedRNNCells`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/StackedRNNCells): Wrapper allowing a stack of RNN cells to behave as a single cell. [`class Subtract`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Subtract): Layer that subtracts two inputs. [`class ThresholdedReLU`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/ThresholdedReLU): Thresholded Rectified Linear Unit. [`class TimeDistributed`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/TimeDistributed): This wrapper allows to apply a layer to every temporal slice of an input. [`class UpSampling1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/UpSampling1D): Upsampling layer for 1D inputs. [`class UpSampling2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/UpSampling2D): Upsampling layer for 2D inputs. [`class UpSampling3D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/UpSampling3D): Upsampling layer for 3D inputs. [`class Wrapper`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/Wrapper): Abstract wrapper base class. [`class ZeroPadding1D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/ZeroPadding1D): Zero-padding layer for 1D input (e.g. temporal sequence). [`class ZeroPadding2D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/ZeroPadding2D): Zero-padding layer for 2D input (e.g. picture). [`class ZeroPadding3D`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/ZeroPadding3D): Zero-padding layer for 3D data (spatial or spatio-temporal). ## Functions [`Input(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/Input): `Input()` is used to instantiate a Keras tensor. [`add(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/add): Functional interface to the `Add` layer. [`average(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/average): Functional interface to the `Average` layer. [`concatenate(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/concatenate): Functional interface to the `Concatenate` layer. [`deserialize(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/deserialize): Instantiates a layer from a config dictionary. [`dot(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/dot): Functional interface to the `Dot` layer. [`maximum(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/maximum): Functional interface to the `Maximum` layer that computes [`minimum(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/minimum): Functional interface to the `Minimum` layer. [`multiply(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/multiply): Functional interface to the `Multiply` layer. [`serialize(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/serialize) [`subtract(...)`](https://www.tensorflow.org/versions/r1.15/api_docs/python/tf/keras/layers/subtract): Functional interface to the `Subtract` layer.