View source on GitHub
|
Layers, based on tf.keras.layers.Layer.
Classes
class GDN: Generalized divisive normalization layer.
class GDNParameter: Nonnegative parameterization as needed for GDN parameters.
class IdentityInitializer: Initialize to the identity kernel with the given shape.
class Parameter: Reparameterized Layer variable.
class RDFTParameter: RDFT reparameterization of a convolution kernel.
class SignalConv1D: 1D convolution layer.
class SignalConv2D: 2D convolution layer.
class SignalConv3D: 3D convolution layer.
class SoftRound: Applies a differentiable approximation of rounding.
class SoftRoundConditionalMean: Conditional mean of inputs given noisy soft rounded values.