LocalResponseNorm#

classtorch.nn.modules.normalization.LocalResponseNorm(size,alpha=0.0001,beta=0.75,k=1.0)[source]#

Applies local response normalization over an input signal.

The input signal is composed of several input planes, where channels occupy the second dimension.Applies normalization across channels.

$$b_c=a_c{(k+\frac{\alpha }{n}\sum _{c^{\mathrm{\prime }}=\max (0,c-n\mathrm{/}2)}^{\min (N-1,c+n\mathrm{/}2)}{a}_{c^{\mathrm{\prime }}}^2)}^{-\beta }$$bc​=ac​​k+nα​c′=max(0,c−n/2)∑min(N−1,c+n/2)​ac′2​​−β

Parameters

• size (int) – amount of neighbouring channels used for normalization

• alpha (float) – multiplicative factor. Default: 0.0001

• beta (float) – exponent. Default: 0.75

• k (float) – additive factor. Default: 1

Shape:

• Input:$(N,C,\ast )$(N,C,∗)

• Output:$(N,C,\ast )$(N,C,∗) (same shape as input)

Examples:

>>>lrn=nn.LocalResponseNorm(2)>>>signal_2d=torch.randn(32,5,24,24)>>>signal_4d=torch.randn(16,5,7,7,7,7)>>>output_2d=lrn(signal_2d)>>>output_4d=lrn(signal_4d)

extra_repr()[source]#

Return the extra representation of the module.

forward(input)[source]#

Runs the forward pass.

Return type

Tensor