Base class for neural network layers.


@interface MPSCNNKernel : MPSKernel


An MPSCNNKernel object consumes one MPSImage object and produces one MPSImage object.

The region overwritten in the destination image is described by the clipRect property. The top left corner of the region consumed (ignoring adjustments for filter size—for example, convolution filter size) is given by the offset property. The size of the region consumed is a function of the size of the clipRect property and any subsampling caused by pixel strides at work (for example, strideInPixelsX/strideInPixelsY in the MPSCNNPooling class). Wherever the offset and clipRect properties would cause an {x,y} pixel address not in the image to be read, the edgeMode property is used to determine what value to read there.

The z or depth component of the offset, origin and size properties indexes which images to use.

  • If the MPSImage object contains only a single image, then these values should be offset.z = 0, clipRect.origin.z = 0, and clipRect.size.depth = 1.

  • If the MPSImage object contains multiple images, then the value of clipRect.size.depth determines the number of images to process. Both the source and destination MPSImage objects must have at least this many images. The value of offset.z refers to the starting image index of the source. Thus, the value of offset.z + clipRect.size.depth must be <=source.numberOfImages. Similarly, the value of clipRect.origin.z determines the starting image index of the destination. Thus, the value of clipRect.origin.z + clipRect.size.depth must be <=destination.numberOfImages.

The destinationFeatureChannelOffset property can be used to control where the kernel will start writing in terms of feature channel dimension. For example, if the destination has 64 channels and thdestinationFeatureChannelOffsete kernel outputs 32 channels, channels 0-31 of the destination will be populated by the kernel (by default). But if you want the kernel to populate channels 32-63 of the destination, you can set the value of destinationFeatureChannelOffset to 32. Suppose you have a source of dimensions w x h x Ni, where N is the number of channels, which goes through a convolution filter C0 which produces the output O0 = w x h x N0 and C1 which produces the output O1 = w x h x N1 followed by concatenation which produces O = w x h x (N0 + N1). You can achieve this by creating an MPSImage object with dimensions w x h x (N0 + N1) and using this as the destination of both convolutions as follows:

  • C0: destinationFeatureChannelOffset = 0, this will output N0 channels starting at channel 0 of destination thus populating [0,N0-1] channels.

  • C1: destinationFeatureChannelOffset = N0, this will output N1 channels starting at channel N0 of destination thus populating [N0,N0+N1-1] channels.


Instance Properties


The position of the destination image's clip rectangle origin, relative to the source image.


A signed coordinate with x, y, and z components.


An optional clip rectangle to use when writing data. Only the pixels in the clip rectangle will be overwritten.


The bounds for a subset of an object's elements.


The number of channels in the destination image to skip before writing output data.


The edge mode to use when texture reads stray off the edge of an image.


The options used to control the edge behavior of an image filter when it reads outside the bounds of a source texture.


The protocol that provides a description of how kernels should pad images.


Inherits From

See Also

Layer Base Classes


A convolution neural network kernel.


The base class for gradient layers.