Apply additive zero-centered Gaussian noise.
This is useful to mitigate overfitting (you could see it as a form of random data augmentation). Gaussian Noise (GS) is a natural choice as corruption process for real valued inputs.
As it is a regularization layer, it is only active at training time.
stddev: Float, standard deviation of the noise distribution.
inputs: Input tensor (of any rank).
training: Python boolean indicating whether the layer should behave in training mode (adding noise) or in inference mode (doing nothing).
Arbitrary. Use the keyword argument
(tuple of integers, does not include the samples axis)
when using this layer as the first layer in a model.
Same shape as input.
__init__( stddev, **kwargs )