DepthToSpace

Description

Rearranges data from depth into spatial data blocks.

Input

One input:

input: tensor input in NCHW format. Must be of type float16 or float.

Output

One output:

output: tensor with shape [N, C/(blocksize * blocksize), H * blocksize, W * blocksize].

Attribute

blocksize: int, size of the block to be moved. This parameter is mandatory.

mode: string, specifying whether the arrangement is depth-column-row or column-row-depth. The default value is DCR.

ONNX Opset Support

Opset v8/v9/v10/v11/v12/v13/v14/v15/v16/v17/v18