tf.contrib.layers.maxout( inputs, num_units, axis=-1, scope=None )
Adds a maxout op from https://arxiv.org/abs/1302.4389
"Maxout Networks" Ian J. Goodfellow, David Warde-Farley, Mehdi Mirza, Aaron Courville, Yoshua Bengio
Usually the operation is performed in the filter/channel dimension. This can also be used after fully-connected layers to reduce number of features.
inputs: Tensor input
num_units: Specifies how many features will remain after maxout in the
axisdimension (usually channel). This must be multiple of number of
axis: The dimension where max pooling will be performed. Default is the last dimension.
scope: Optional scope for variable_scope.
Tensor representing the results of the pooling operation.
ValueError: if num_units is not multiple of number of features.