SpaceToBatch

Versioned name: SpaceToBatch-2

Category: Data movement

Short description: The SpaceToBatch operation divides "spatial" dimensions [1, ..., N - 1] of the data input into a grid of blocks of shape block_shape, and interleaves these blocks with the batch dimension (0) such that in the output, the spatial dimensions [1, ..., N - 1] correspond to the position within the grid, and the batch dimension combines both the position within a spatial block and the original batch position. Prior to division into blocks, the spatial dimensions of the input are optionally zero padded according to pads_begin and pads_end.

Detailed description:

The SpaceToBatch operation is similar to the TensorFlow* operation SpaceToBatchND

The operation is equivalent to the following transformation of the input tensor data of shape [batch, D_1, D_2 ... D_{N - 1}] and block_shape, pads_begin, pads_end of shapes [N] to Y output tensor.

Zero-pad the start and end of dimensions [D_0, ..., D_{N - 1}] of the input according to `pads_begin` and `pads_end`:
note: P_0 for batch dimension is expected to be 0 (no-padding).
x = [batch + P_0, D_1 + P_1, D_2 + P_2, ..., D_{N - 1} + P_{N - 1}], where P_i = pads_begin[i] + pads_end[i]

note: B_0 for batch is ignored.
x' = reshape(x, [batch, (D_1 + P_1) / B_1, B_1, (D_2 + P_2) / B_2, B_2, ..., (D_{N - 1} + P_{N - 1}) / B_{N - 1}, B_{N - 1}]), where B_i = block_shape[i]

x'' = transpose(x',  [2, 4, ..., (N - 1) + (N - 1), 0, 1, 3, ..., N + (N - 1)])

y = reshape(x'', [batch * B_1 * ... * B_{N - 1}, (D_1 + P_1) / B_1, (D_2 + P_2) / B_2, ... , (D_{N - 1} + P_{N - 1}) / B_{N - 1}])

Attributes

No attributes available.

Inputs

• 1: data - input N-D tensor [batch, D_1, D_2 ... D_{N - 1}] of T1 type with rank >= 2. Required.
• 2: block_shape - input 1-D tensor of T2 type with shape [N] that is equal to the size of data input shape. All values must be >= 1. block_shape[0] is expected to be 1. Required.
• 3: pads_begin - input 1-D tensor of T2 type with shape [N] that is equal to the size of data input shape. All values must be non-negative. pads_begin specifies the padding for the beginning along each axis of data input . It is required that block_shape[i] divides data_shape[i] + pads_begin[i] + pads_end[i]. pads_begin[0] is expected to be 0. Required.
• 4: pads_end - input 1-D tensor of T2 type with shape [N] that is equal to the size of data input shape. All values must be non-negative. pads_end specifies the padding for the ending along each axis of data input. It is required that block_shape[i] divides data_shape[i] + pads_begin[i] + pads_end[i]. pads_end[0] is expected to be 0. Required.

Outputs

• 1: N-D tensor with shape [batch * block_shape[0] * block_shape[1] * ... * block_shape[N - 1], (pads_begin[1] + D_1 + pads_end[1]) / block_shape[1], (pads_begin[2] + D_2 + pads_end[2]) / block_shape[2], ..., (pads_begin[N - 1] + D_{N - 1} + pads_end[N - 1]) / block_shape[N - 1] of the same type as data input.

Types

• T1: any supported type.
• T2: any supported integer type.

Example

<layer type="SpaceToBatch" ...>
    <input>
        <port id="0">       <!-- data -->
            <dim>2</dim>    <!-- batch -->
            <dim>6</dim>    <!-- spatial dimension 1 -->
            <dim>10</dim>   <!-- spatial dimension 2 -->
            <dim>3</dim>    <!-- spatial dimension 3 -->
            <dim>3</dim>    <!-- spatial dimension 4 -->
        </port> 
        <port id="1">       <!-- block_shape value: [1, 2, 4, 3, 1] -->
            <dim>5</dim>  
        </port>
        <port id="2">       <!-- pads_begin value: [0, 0, 1, 0, 0] -->
            <dim>5</dim>
        </port>
        <port id="3">       <!-- pads_end value: [0, 0, 1, 0, 0] -->
            <dim>5</dim>
        </port>
    </input>
    <output>
        <port id="3">
            <dim>48</dim>   <!-- data.shape[0] * block_shape.shape[0] * block_shape.shape[1] *... * block_shape.shape[4] -->
            <dim>3</dim>    <!-- (data.shape[1] + pads_begin[1] + pads_end[1]) / block_shape.shape[1]  -->
            <dim>3</dim>    <!-- (data.shape[2] + pads_begin[2] + pads_end[2]) / block_shape.shape[2] -->
            <dim>1</dim>    <!-- (data.shape[3] + pads_begin[3] + pads_end[3]) / block_shape.shape[3] -->
            <dim>3</dim>    <!-- (data.shape[4] + pads_begin[4] + pads_end[4]) / block_shape.shape[4] -->
        </port>
    </output>
</layer>