I am building an autoencoder for 3D images and would like to use Depthwise convolutions. For the encoder, I found an implementation of a depthwise 3D convolutional layer (DepthwiseConv3D).
For the decoder, I need a DepthwiseConv3DTranspose layer.
I did not find an implementation of this anywhere. I read that "a deconvolution layer performs also convolution".
This made me think: is it possible to choose parameters so that DepthwiseConv3D performs a transposed convolution?