2D Transposed Convolution Forward Layer

The forward two-dimensional (2D) transposed convolution layer computes the tensor Y by applying a set of nKernels 2D kernels K of size m 3 x m 4 to the input tensor X.

Problem Statement

The problem is to compute the four-dimensional tensor of values YR n 1 x nKernels x n 3 x n 4 such that:


For the notations in this formula, refer to 2D Convolution Backward Layer.

The computation flow in the forward 2D transposed convolution layer is identical to the computation of the gradient in the 2D convolution backward layer, except the following notation changes:

2D Convolution Backward Layer

2D Transposed Convolution Forward Layer

Input gradient tensor G

Input tensor X

Gradient tensor Z

Result tensor Y


l 2

n 2


For more complete information about compiler optimizations, see our Optimization Notice.
Select sticky button color: 
Orange (only for download buttons)