Class Dilation2d<T extends TNumber>

java.lang.Object
org.tensorflow.op.RawOp
org.tensorflow.op.nn.Dilation2d<T>
All Implemented Interfaces:
Shaped, Op, Operand<T>
@Operator(group="nn") public final class Dilation2d<T extends TNumber> extends RawOp implements Operand<T>
Computes the grayscale dilation of 4-D input and 3-D filter tensors. The input tensor has shape [batch, in_height, in_width, depth] and the filter tensor has shape [filter_height, filter_width, depth], i.e., each input channel is processed independently of the others with its own structuring function. The output tensor has shape [batch, out_height, out_width, depth]. The spatial dimensions of the output tensor depend on the padding algorithm. We currently only support the default "NHWC" data_format.

In detail, the grayscale morphological 2-D dilation is the max-sum correlation (for consistency with conv2d, we use unmirrored filters): 

output[b, y, x, c] =
   max_{dy, dx} input[b,
                      strides[1] * y + rates[1] * dy,
                      strides[2] * x + rates[2] * dx,
                      c] +
                filter[dy, dx, c]

Max-pooling is a special case when the filter has size equal to the pooling kernel size and contains all zeros.

Note on duality: The dilation of input by the filter is equal to the negation of the erosion of -input by the reflected filter.

  • Field Details

  • Constructor Details

    • Dilation2d

      public Dilation2d(Operation operation)

  • Method Details

    • create

      @Endpoint(describeByClass=true) public static <T extends TNumber> Dilation2d<T> create(Scope scope, Operand<T> input, Operand<T> filter, List<Long> strides, List<Long> rates, String padding)
      Factory method to create a class wrapping a new Dilation2D operation.
      Type Parameters:
      T - data type for Dilation2D output and operands
      Parameters:
      scope - current scope
      input - 4-D with shape [batch, in_height, in_width, depth].
      filter - 3-D with shape [filter_height, filter_width, depth].
      strides - The stride of the sliding window for each dimension of the input tensor. Must be: [1, stride_height, stride_width, 1].
      rates - The input stride for atrous morphological dilation. Must be: [1, rate_height, rate_width, 1].
      padding - The type of padding algorithm to use.
      Returns:
      a new instance of Dilation2d

    • output

      public Output<T> output()
      Gets output. 4-D with shape [batch, out_height, out_width, depth].
      Returns:
      output.

    • asOutput

      public Output<T> asOutput()
      Description copied from interface: Operand
      Returns the symbolic handle of the tensor.

      Inputs to TensorFlow operations are outputs of another TensorFlow operation. This method is used to obtain a symbolic handle that represents the computation of the input.

      Specified by:
      asOutput in interface Operand<T extends TNumber>
      See Also: