AveragePooling2D

classin packageklyn.math.nn.layers

public class AveragePooling2D<T extends AbstractFloat = Float> extends Pooling2DLayer<T>

AveragePooling2D

Mean over each spatial window, excluding padded positions from its divisor. Inputs and outputs use [batch, height, width, channels] order. Supports overlapping windows and valid/same padding, with complete input gradients.

Example

layer = AveragePooling2D<Float>(2)

Constructors

Modifier and Type	Member	Description
public	AveragePooling2D `AveragePooling2D(poolSize as Long = 2l, strides as Long = -1l, padding as String = "valid")`	Configures independent window reductions with no trainable parameters.
public	AveragePooling2D `AveragePooling2D(poolSize as Long, strides as Tuple, padding as String = "valid")`	Configures independent window reductions with no trainable parameters.
public	AveragePooling2D `AveragePooling2D(poolSize as Tuple, strides as Long, padding as String = "valid")`	Configures independent window reductions with no trainable parameters.
public	AveragePooling2D `AveragePooling2D(poolSize as Tuple, strides as Tuple = null, padding as String = "valid")`	Configures independent window reductions with no trainable parameters.