3D Deconvolution CPU support

[aulombard]

Description

Until today mxnet does not support 3D Deconvolution (Transposed convolution) layers on CPU (even with mkldnn support) as it is only implemented for GPU with Cudnn support. Is there any plan in the future to support it knowing that mkldnn is supposed to support it and knowing that most of other frameworks support this feature. This layer is used a lot in UNet like networks, which is a very well known architecture used in computer vision problems.

References

[github-actions]

Welcome to Apache MXNet (incubating)! We are on a mission to democratize AI, and we are glad that you are contributing to it by opening this issue.
Please make sure to include all the relevant context, and one of the @apache/mxnet-committers will be here shortly.
If you are interested in contributing to our project, let us know! Also, be sure to check out our guide on contributing to MXNet and our development guides wiki.

[anko-intel]

@mxnet-label-bot add [MKLDNN]