local_response_normalization

dragon.vm.tensorflow.nn.local_response_normalization(
  input,
  depth_radius=5,
  bias=1.0,
  alpha=1.0,
  beta=0.5,
  data_format='NHWC',
  name=None
)[source]

Apply the local response normalization. [Krizhevsky et.al, 2012].

The normalization is defined as:

\[out_{i} = x_{i}\left(k + \frac{\alpha}{n} \sum_{j=\max(0, i-n/2)}^{\min(N-1,i+n/2)}x_{j}^2 \right)^{-\beta} \]
Parameters:
  • input (dragon.Tensor) – The input tensor.
  • depth_radius (int, optional, default=5) – The number of neighbouring channels to sum over.
  • bias (float, optional, default=1.) – The bias constant \(k\).
  • alpha (float, optional, default=1.) – The scale value \(\alpha\).
  • beta (float, optional, default=0.5) – The exponent value \(\beta\).
  • data_format ({'NCHW', 'NHWC'}, optional) – The optional data format.
  • name (str, optional) – The operation name.
Returns:

dragon.Tensor – The output tensor.